S21 .A7) 63.06(73) U£S. DEPARTMENT OF AGRICULTURE Department Bulletin 1250-13^9 ' . \ ' .Jt' i i^$4i£g S ' JI5 Rt wl'3 *■ m: m . mm Mi c* urn ,jv 1 m il I m i, ty-S w (SfCpyLv^.i- mui* i5Mr •%, ■ 1 ' I /■*£% Twk “ .< %w ■Pfcffir, % tf ii. 16* .# > tL? V * mjj Nurl Mr?' I. T: V Jfc :.< •- ^i^afcv. . rV^W V* Ikit %.jSr 'tfh WJ-t 'mi-jia- . * i Hi wXi'W* FOR THE PEOPLE FOR$$V cation LIBRARY OF THE AMERICAN MUSEUM OF NATURAL HISTORY U.S. DEPARTMEN T OF AGRICULTURE Department Bulletin 1250-1349 Only the numbers listed below have been retained : 1267 1324 1268 1328 1273 1332 1303 1336 1313 1339 1320 1346 13^9 U. S. DEPARTMENT OF AGRICULTURE OFFICE OF INFORMATION DIVISION OF PUBLICATIONS DEPARTMENT BULLETINS Nos. 1251-1275 WITH CONTENTS PREPARED IN THE INDEXING SECTION UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON 1928 CONTENTS Department Bulletin No. 1251. — Effect of Winter Rations on Pasture Gains of 2-Year-Old Steers. I. Winter Rations and Their Influence on Pasture Gains of 2-Year-Old Steers. II. The Cost of Rations for Wintering 2-Year-Old Steers: I. Winter rations and their influence on pasture gains of 2-year-old steers 1 The Appalachian region and its problems 1 Objects and plan of the work 2 Kind of steers used 4 Feeds used 4 Character of pasture 4 Method of feeding and handling the steers 5 Quantity of feed consumed 5 Gains during winter and summer 6 Diagrams of gains and losses 7 Correlations 15 II. Cost of rations for wintering 2-year-old steers 16 Cost per pound of gain 17 Shrinkage in transit and dressing percentage 23 Conclusions 24 Department Bulletin No. 1252.— Prune and Cherry Brown-Rot Investigations in the Pacific Northwest: Introduction 1 Blossom infection 2 Apothecia ! 3 Time relation of apothecia, blossoms, and blossom infection 3 Relation of plowing, cultivation, and character of soil 3 Fruit infection 4 Spraying experiments on prunes 4 Spraying experiments in 1915 4 Spraying experiments in 1916 6 Spraying experiments in 1917 7 Spraying experiments in 1918 8 Spraying experiments in 1919 9 Summary of spraying experiments on prunes 10 Recommendations for the control of brown-rot of prunes 11 Spraying experiments on cherries 12 Spraying experiments in 1915 12 Spraying experiments in 1916 13 Spraying experiments in 1917 : 13 Spraying experiments in 1918 15 Spraying experiments in 1919 16 Summary of spraying experiments on cherries 16 Recommendations for the control of brown-rot of cherries 17 Preparation of sprays 17 Self-boiled lime sulphur 17 Bordeaux mixture 18 Lime-sulphur solution 19 Rosin-fishoil soap 19 Casein 20 Spraying schedule 20 Summary 21 79409E— 28 3 4 DEPARTMENT OF AGRICULTURE BULS. 1251-1275 Page Department Bulletin No. 1253. — Diseases of Apples on the Market: Introduction 1 Tabulation of data 3 Analysis of data 5 Box crop: Disease by years 5 Barrel crop: Disease by years 6 Box crop: Disease by months 7 Barrel crop: Disease by months 9 Box crop: Disease by varieties 10 Barrel crop: Disease by varieties 11 Box crop: Disease by States 12 Barrel crop: Disease by States 13 Box and barrel crop: Disease by inspection offices 13 Disease by crops: Barrel and box 17 Scald by crops 18 Relation of rots, scald, internal breakdown, and water core to size and grade 22 Freezing injury 23 Summary 24 Department Bulletin No. 1254. — Farm Motor Truck Operation in the New England and Central Atlantic States: Collection of data 1 Summary 2 Number and location of truck owners reporting 3 Types of farms on which trucks are owned 3 Distance to market 4 Size of trucks 5 Age of trucks 7 Are these trucks profitable investments? 7 Advantages and disadvantages of motor trucks 7 The best size of truck 9 Change of markets 10 Road hauling with trucks 11 Road hauling for which trucks are not used 13 Effect of different kinds of roads on use of motor trucks 14 Hauling on the farm with trucks 15 Custom hauling ^ 16 Annual use of trucks 16 Cost of operation . 18 Cost of hauling with motor trucks 23 Reliability 24 Saving of hired help 25 Displacement of horses 26 Farms on which both trucks and tractors are owned 27 Department Bulletin No. 1255. — Inheritance of Composition in Fruit Through Vegetative Propagation. — Bud Variants of Eureka and Lisbon Lemons: Introduction 1 The problem 2 Experimental procedure: Sampling 3 Significance of determinations made 3 Methods of analysis 4 Results 5 Discussion of results 14 Variability in citrus fruits 14 Differences in composition of fruit from the same tree and from trees of the same strain 14 Differences in composition of fruit from trees of different strains. 15 Summary 18 Literature cited 18 CONTENTS 5 Page Department Bulletin No. 1256. — Tobacco Diseases and Their Control: Introduction 1 The nature of plant diseases 3 Losses from disease 3 The plant bed as a source of infection 4 Plant-bed sanitation 4 The sterilization of soil for plant beds 5 Stem or stalk diseases 7 Root diseases 18 Leaf diseases 25 Injuries due to physical agencies 41 Unimportant or rare diseases not otherwise classified 41 Damage in curing and fermentation 43 Diseases of tobacco in foreign countries 48 Bibliography 50 Department Bulletin No. 1257.— Land Reclamation Policies in the United States: Past land reclamation policies 3 Federal policies 3 Swamp land acts 3 Relation of the homestead act to reclamation policies 3 Act of 1866 4 The desert land act 4 Irrigation survey 5 The Carey Act ■ 6 Reclamation act 7 Irrigation district act 13 State land reclamation policies 14 General considerations 14 Irrigation districts 14 California State land settlement 20 Proposed Federal and State cooperation 23 Drainage reclamation 24 State drainage policies 24 Summary of acreage reclaimed 27 The future of reclamation 34 Conclusions 40 Department Bulletin No. 1258. — Farm Organization and Man- agement in Clinton County, Indiana. — A Business Analysis of 100 Farms, in Forest and Johnson Townships, for Eight Years — 1910, and 1913 to 1919, Inclusive: Description of area 2 Definition of terms 3 Organization and management of 100 farms 4 Livestock production 20 Work animals < 25 Man labor 26 Farm capital 28 Farm receipts 30 Farm expenses 34 Farm earnings 35 Quantities of products sold 42 Prices received for products 42 Acres, crop yields per acre, etc 43 Farm tenure 50 Summary 66 6 DEPARTMENT OF AGRICULTURE BULS. 1251-1275 Pag* Department Bulletin No. 1259. — Standard Specifications for Steel Highway Bridges.- — Adopted by the American Association of State Highway Officials and Approved by the Secretary of Agriculture for Use in Federal-Aid Road Work:’ Introduction ii Division I. Materials 1 Structural, rivet, and eyebar steel 1 E}?ebars 2 Steel forgings 3 Wrought iron 3 Steel castings 3 Gray-iron castings 4 Malleable castings 4 Phosphor-bronze 4 Division II. General construction 4 Steel structures 4 Painting steel structures 15 Division III. Design 18 General features 18 Loads 20 Unit stresses 25 Distribution of loads 26 Structural steel design 27 Index 46 Department Bulletin No. 1260. — Sorghum Experiments on the Great Plains: The sorghum belt 1 Climatic features 2 Soils . 7 Native vegetation 9 Sorghum groups and varieties 10 Experimental methods 11 Preservation of soil uniformity 11 Plat technique 11 Seeding methods 11 Harvesting methods 12 Methods of obtaining data 12 Varietal experiments 13 Hays, Ivans 13 Chillicothe, Tex 23 Amarillo, Tex 32 Woodward, Okla 38 Lawton, Okla 40 Dalhart, Tex 41 Big Spring, Tex 43 Tucumcari, N. Mex 44 Northern Great Plains 46 Summary of varietal experiments 47 Cultural experiments ; 57 Date of seeding 57 Rate of seeding in rows 72 Rate of seeding in close drills or broadcast 83 Time of cutting sorghum for hay 85 Literature cited 88 Department Bulletin No. 1261. — Operating Methods and Expense of Cooperative Citrus-Fruit Marketing Agencies: Organization of the exchange system 1 Harvesting operations and expense 3 Packing-house operation 6 Management : 6 Special operating problems 7 Careful handling 7 Standardized grades 8 Pools and payments to growers 9 CONTENTS 7 Page Department Bulletin No. 1261.— Operating Methods and Expense of Cooperative Citrus-Fruit Marketing Agencies — Continued. Packing-house operation — Continued. Regular packing-house operations 10 Grading, sizing, and packing oranges 11 Grading and packing lemons 12 Loading orange and lemon shipments 14 Packing-house labor 14 Packing-house expense 14 Average expense 18 Increases in expense 21 Variations in expense 23 Growers’ receipts 28 Expense of distribution 30 District exchange expense 31 Expense of the central exchange 31 Expense of transportation 32 Wholesale and retail margins 33 Department Bulletin No. 1262. — Effect of Kiln Drying, Steam- ing, and Air Seasoning on Certain Fungi in Wood: Introduction 1 Previous work 2 Material used in this study 3 Methods of study 5 Cultural methods 5 Kiln-dr ying and steaming experiments 6 Brief summary of the test runs 10 Air-seasoning experiments 14 Revival of fungi in wood after air drying 15 Review of the results 17 Summary il 19 Literature cited 20 Department Bulletin No. 1263. — Relative Resistance of Tree Seedlings to Excessive Heat: The problem 1 Literature available 2 Description of experiments 5 Preliminary tests 5 Plan of tests in 1922 5 Results in moist air 30 days after sowing 7 Results in dry air 46, 64, and 90 to 92 days after sowing 9 Influence of age of seedlings 12 Temperature scale for each species 12 Conclusions 13 Literature cited 16 Department Bulletin No. 1264. — Forest Planting in the Inter- mountain Region: Introduction 1 Seed collection 2 Yield from cones 3 Cost of seed 3 Nursery practice 4 Nursery operations 5 Protection from diseases and injuries 6 Distribution of planting stock 9 Field planting 10 Planting sites and native timber types 10 Methods of planting 11 Number of plants per acre 12 Results of field planting 13 Causes of loss and failure and methods of prevention 39 Planting costs 45 Future of artificial forestation in intermountain region 45 Summary 47 Appendix 49 Literature cited 56 8 DEPARTMENT OF AGRICULTURE BULS. 1251-1275 Page Department Bulletin No. 1265. — Scalding, Precooking, and Chill- ing as Preliminary Canning Operations: Introduction 1 Apparatus and methods used 3 Experiments with specific food materials 4 General discussion 33 General summary 37 Literature cited 39 A list of pertinent nontechnical literature 43 Department Bulletin No. 1266. — Agricultural Cooperation in Denmark: Yearly average exchange rates on Danish krone ii Introduction 1 History and development of Danish agriculture 4 Danish agriculture to-day 7 Principles observed in Danish cooperation 9 Dairy industry and cooperation 14 Bacon industry and cooperation 31 Egg industry and cooperation 46 Cooperative cattle export associations 54 Cooperative buying 56 Cooperative breeding associations 74 Agricultural credit 81 Miscellaneous cooperative organizations : 87 Department Bulletin No. 1267.— The Plough-Headed Cornstalk- Beetle: Introduction 1 Economic history 2 Distribution 3 Life history 4 General account 4 Egg 5 Larva 9 Prepupa 16 Pupa 17 Adult 18 Species likely to be mistaken for Euetheola rugiceps 26 Ligyrus gibbosus (De Geer) 26 Dyscinetus trachypygus (Burm.) 28 Cyclocephala spp 29 Phyllophaga spp 29 Natural enemies 29 Control measures 31 Elimination of waste lands and old pastures 31 Pasturing with hogs 31 Early planting 32 Change of rotation 32 Fertilizers 32 Hand picking 32 Late summer plowing 32 Summary of control measures 32 Literature cited 33 Department Bulletin No. 1268. — Returns prom Banded Birds, 1920 to 1923: Introduction 1 Organized activities in bird banding 1 Regional banding associations 3 Returns reported to the Biological Survey 4 Explanation of tables 5 Tables of returns 6 Index 54 CONTENTS 9 Page Department Bulletin No. 1269. — Relation of Land Tenure to Plantation Organization: Area and extent of the plantation system 2 Characteristics of the plantation system 8 Organization and management 11 Plantation labor 19 Wage labor 23 Cropper labor 29 Tenant labor 32 Relations of laborers and tenants to plantation operators and land- lords 38 Renting arrangements 38 Labor supervision 42 Labor movements and occupancy 44 Selection of enterprises, and diversification 52 Credit 60 Marketing 65 Conclusions 67 Department Bulletin No. 1270. — The Production of Narcissus Bulbs: Explanation of terms 1 Securing stock for planting 2 How bedding and forced bulbs should be handled 4 Preparation of the bulbs for planting 5 Preparation of the soil 5 Planting 5 Time of planting 6 Autumn treatment of the beds 6 Cultivation 7 Mulching 7 Spring work on the beds 7 Second year of the biennial crop 8 Roguing 8 Disposition of the rogues 8 Digging 8 Daffodil and tulip digging compared 9 Removing loose soil from the bulbs 9 Storage 10 Curing 11 Changes in the bulbs as they dry 11 Breaking the bulbs apart 12 Sizers and sizing 13 Drainage and soil percolation 14 Use of lime 15 Fertilizers in commercial culture 15 Daffodils a biennial crop 16 Culture for cut flowers 16 Harvesting flowers 17 Removing the faded flowers 17 Appearance of the flower spike 17 Difference in cost of varieties 18 Special items 18 Enemies 23 Where narcissus bulbs are grown «_ 24 Relative use of varieties 25 Breeding daffodils 25 Naturalizing varieties 26 Yields 26 Conditions of daffodil culture 28 Narcissus varieties and their classification 29 Recommendations 30 10 DEPARTMENT OF AGRICULTURE BULS. 1251-1275 Page Department Bulletin No. 1271. — A Study of Farm Organization in Southwestern Minnesota: Description and history of agriculture of the area 2 Purpose of historical study 4 Settlement and agricultural development 5 Changes in acreages of the important crops 6 Changes in the numbers and kinds of livestock 8 Study of present-day agriculture in Cottonwood and Jackson Counties 11 Unit requirements of labor and materials for crops 13 Seed-bed preparation 14 Corn 17 Oats 24 Barley 29 Rye 32 Flax 35 Tame hay 36 Wild hav 39 Alfalfa- 1 41 Use of unit requirement and labor distribution data in planning a cropping system 44 Unit requirements of labor and materials for livestock 45 Work horses 46 Colts-.- 48 Dairy cows 50 Young dairy cattle 53 Mixed cattle 55 Hogs 58 Sheep 61 Poultry 63 Miscellaneous labor requirements and their relation to the crop and and livestock labor 66 Manure hauling 66 Miscellaneous crop labor 67 Miscellaneous livestock labor 67 Maintenance labor 67 The place of the crop, livestock, and miscellaneous labor in the labor program of the farm 70 Day-to-day management of labor 75 Exchange labor 78 The principles of the choice and adjustment of crop and livestock enterprises 79 Factors affecting choice and adjustment of enterprises 79 General plan of application of the principles of choice and adjustment of enterprises 82 Summary 99 Department Bulletin No. 1272. — Values of Various New Feeds for Dairy Cows: Plan of the experimental work 1 Fish meal compared with cottonseed meal 2 Peanut feed compared with cottonseed meal 3 Potato meal compared with corn meal 5 Velvet-bean meal compared with cottonseed meal 6 Sweet-potato meal compared with corn meal 7 Potato silage compared with corn silage 7 Apple-pectin pulp compared with beet pulp 8 Hydrolyzed sawdust compared with corn meal 9 Cane molasses as a supplementary feed 12 Summary 14 CONTENTS 11 Department Bulletin No. 1273. — The Bud Moth: Introduction 1 Historical 1 Synonymy 2 Common name 3 Food plants 3 Distribution 4 Means of dissemination 5 Economic importance 5 Other species of bud moth 5 Descriptions of stages of the bud moth 6 Seasonal history and habits 8 Natural enemies 14 Control 16 Summary 19 Literature cited 19 Department Bulletin No. 1274. — Cockleburs (Species of Xanthium) as Poisonous Plants: Purpose and scope 1 Historical summary 2 The cocklebur plant 5 Experimental work 6 Typical case of pig 18 6 Discussion and general conclusions 7 Symptoms in pigs 7 Symptoms in sheep and cattle 12 Symptoms in chickens 12 Autopsy findings 12 Microscopic changes in tissues 13 Toxic and lethal dosage 15 Time from feeding to appearance of symptoms 16 Duration of sickness 16 Effect of continued feeding 17 Animals poisoned by cocklebur 18 Part of plant poisonous 18 Mechanical injury by burs 19 Toxicity of dried plant 20 Remedies 21 Summary 22 Bibliography 22 Department Bulletin No. 1275. — Varietal Susceptibility of Oats to Loose and Covered Smuts: Introduction 1 Review of literature 1 Characteristics of the oat smuts 5 Field experiments 6 Experimental procedure and results 6 Discussion of results 29 Greenhouse experiments 34 Summary 36 Literature cited 38 UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1267 Washington, D. C. Y October 14, 1924 THE ROUGH-HEADED CORN STALK-BEETLE1 By W. J. Phillips, Entomologist, and Henry Pox,2 Entomological Assistant, Cereal and Forage Insect Investigations, Bureau of Entomology CONTENT'S Introduction Economic history Distribution Life history General account Egg Larva Prepupa Pupa Adult Species likely to be mistaken for Euetheola rugiceps Ligyrus gibbosus (De Geer) Dyscinetus trachypygus ( Burm. ) _ Cyclocephala spp Page 1 2 3 4 4 5 9 16 17 18 26 26 28 29 Species likely to- be mistaken, for Euetheola rugiceps — Continued. Phyllophaga spp- Natural enemies Control measures Elimination of waste lands and old pastures Pasturing with hogs Early planting Change of rotation Fertilizers Hand picking : Late summer plowing Summary of control measures Literature cited Page 29 29 31 31 31 32 32 32 32 32 32 33 INTRODUCTION The ravages of ( Ligyrus ) Euetheola rugiceps (Lee.) were first brought to the attention of the writers in 1914 ( 9 , p. 3). 3 Dr. J. M. Gouldin, of Tappahannock, Essex County, Va,, in a letter dated June 26, 1914, stated that some farmers lost nearly their entire corn crop. The late Prof. F. M. Webster, then in charge of Cereal and Forage Insect Investigations, instructed the senior writer to make a personal survey of the situation. This was done early in July, 1914, and showed that serious damage (PI I, A) had occurred on several hundred acres of corn in the vicinity of Tappahannock, Va. At that time the beetles had practically ceased their activities, but specimens were sent to Charlottesville, Va,, for life-history studies. Since little was known of the habits of this pest or the means of control, the problem of determining these points was assigned to the Charlottesville laboratory, with the senior writer in charge. The breeding records obtained from material secured in 1914 were disappointing, and since the locality of the outbreak was rather 1 Euetheola rugiceps (Lee.) ; order Coleoptera, family Scarabaeidae. 2 Resigned August 31, 1918. 8 Reference is made by number ( italic ) to " Literature cited,” p. 33. 1 94051°— 24- 2 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE inaccessible, temporary headquarters were established in the heart of the infested district, in order to study the problem at first hand. In the spring of 1915 the junior writer was assigned to Tappahan- nock, where he remained until October. In the meantime detailed studies were being conducted in breeding cages at Charlottesville as a check on the work at Tappahannock. The junior writer re- turned to Tappahannock in 1916 to complete the data on the life history and to start field experiments for controlling the pest-, the results of which have already been published in brie! (9) . In this publication it was termed the “ rough-headed corn stalk-beetle,” a translation of the specific name which seems more appropriate than its earlier name of “ sugar-cane beetle,” since the insect has been receiving constantly growing emphasis as a corn pest. ECONOMIC HISTORY Euetheola rugiceps (PI. II, A) was first named and described by John Le Conte (8) in 1856 from specimens obtained in Georgia, and was for a number of years thereafter considered a rather rare southern insect. Riley (10) and Comstock (A) published the first records of the depredations of this pest in 1880, when it first attracted their attention as a sugar-cane insect on the plantations in Louisiana. Comstock stated that the planters in the infested district claimed to have known the pest and had recognized it as a serious menace to sugar cane for a period of about 20 years preceding the outbreak in the seventies. According to the planters, the first serious outbreak occurred about 1855 or 1856; the next destructive one was in 1875. During 1875 and the two succeeding years the depredations on the sugar-cane plantations caused serious alarm, but there seems to have been a decrease in the activities of the insect in 1879. In 1880 the beetles reappeared and inflicted serious loss. This outbreak was re- ported by Comstock (3). Although both Riley and Comstock inci- dentally recorded the insect as injuring corn, it was considered pri- marily a sugar-cane pest and received the vernacular name of “ su- gar-cane beetle,” and by this term it has been designated in the literature until recently. L. O. Howard (7) in 1888 was the first to recognize E. rugiceps as a. corn pest, publishing in that year an ac- count of its depredations to corn in North Carolina and Mississippi. About the same time F. M. Webster (14) made similar observations in Arkansas and Louisiana. In 1895 Weed (15, 16) reported losses to corn growers in Mississippi, but, through what was evidently an error in identification, attributed the damage to Ligyrus gibbosus which it is now believed never injures corn. Since then depredations by this pest have been reported at rather frequent intervals by an in- creasing number of investigators, among whom may be mentioned Titus (13), Garman (6), and Sherman (11), the last-mentioned au- thor especially having published an interesting account of the beetle and its work in North Carolina. The earliest record of injury to corn in Virginia was in 1913, when several farmers reported injury in the “ tidewater ” section of the State. In the following year the depredations were most severe. THE ROUGH-HEADED CORN STALK-BEETLE 3 DISTRIBUTION The known distribution of Euetheola rugiceps in the United States is shown in Figure 1. The data upon which the map is based were obtained from the literature, from hitherto unpublished field records and correspondence of the bureau, and from personal correspondence with a number of museum and experiment station entomologists.4 Euetheola rugiceps is recorded from all the Southern States lying south of the latitude of Washington, D. C., with the exception of Florida and Oklahoma. Judging by the erratic manner in which the species has been observed to occur in the infested sections of Virginia, it would be inadvisable to draw final conclusions regarding the limits of distribution from the evidence at present available. In Virginia the species was found only in that part of the coastal plain which lies between the Potomac and James Rivers, apparently preferring low, moist, poorly drained soils. Even within the area thus restricted, the spe- cies appears at present to be of very unequal distribu- tion, being abun- dant in certain lo- calities and rare or absent in others. It may be found s wanning in certain fields, utterly de- stroying the corn crop, while other fields of the same general type less than a mile away appear to be uninfested. Doubt- less there are unknown factors which influence and limit the spread of this species. The following is a list of localities, arranged according to States, from which there are records of the occurrence of the species. Fig. 1.- -Map showing distribution of Euetheola rugiceps in the United States Alabama. — Birmingham, Catherine, Carollton, Cleveland, Eutaw, Hampden, Hartsells, Mobile, Sprott. Arkansas (by counties). — Ashley, Bradley, Clark, Crawford, Cross, Hot Springs, Howard, Jackson, Jefferson, Lincoln, Lonoke, Monroe, Nevada, Perry, Pope, Pulaski, St. Francis. Georgia. — Bainbridge, Canton, Dalton, Macon. Kentucky. — Guthrie and Hartford. Louisiana. — Atchafalaya River, Baldwin, Baton Rouge, Berwick, Breaux Bridge, Broussard, Castille, Clinton, Church Point, Crowley, Donaldsonville, Franklin, Hester, Koran, La Fayette, Mer Rouge, Mill Haven, Monroe, Mound, Morgan City, New Iberia, New Orleans, Oak Grove, Plaquemine, Rayne, Scott, St. James, St. Joseph, Tensas Parish, Ville Platte. * Those who furnished valuable data in this connection include Franklin Sherman, North Carolina Department of Agriculture; .1. R. Watson, Florida Agricultural Experiment Sta- tion; A. F. Conradi, Clem son College, S. C. ; W. V. Reed, Georgia State Board of Agricul- ture; W. E. Hinds, Alabama Agricultural Experiment Station; J. ,T. Davis and George G. Ainslie, of the Bureau of Entomology; S. J. Hunter, of (he University of Kansas; W. J. Holland, Carnegie Museum; Charles Schaffer, Brooklyn Institute of Arts and Sciences; O. G. Becker, Arkansas Agricultural Experiment Station ; H. Garman, Kentucky Agri- cultural Experiment Station. 4 DEPARTMENT BULLETIN 126*7, U. S. DEPT. OF AGRICULTURE Mississippi. — Agricultural College, Brookhaven, Canton, Durant, Greenwood, Gulfport, Kosciusko, Natchez, Ocean Springs, Winona. North Carolina. — Bostic, Gastonia, Greenville, Monroe, Mount Pleasant, Pan- tego. South Carolina. — Cheraw, Union. Tennessee. — Clarksville, Greeneville, Milan, Savannah, Sevierville. Texas. — Austin, Beaumont, Fedor, Galveston, Jackson County, New Braun- fels, Port Arthur, Victoria. Virginia. — Achilles, Coles Point, Kinsale, Naxera, Odd, Sharps, Tappa- hannock. LIFE HISTORY GENERAL ACCOUNT’ Euetheola ru.giceps hibernates in the soil as an adult in or near its normal feeding grounds. It reappears with warm weather, which in Virginia is in late April or early May. At Tappahannock, Va., the earliest dates on which the beetles were found abroad ' were April 23, 1915, and May 1, 1916. The exact time of their appearance is unquestionably determined by the prevailing weather conditions, being accelerated by high temperatures and retarded by low ones. Thus far the beetles have been found flying only at night, when they are frequently attracted to lights, but it is not an uncommon occur- rence to find them crawling upon the surface of the ground in day- light. The adults begin to feed as soon as they leave their hibernating quarters. Their normal food evidently consists of certain grasses, particularly those belonging to the genus Paspalum, but should these plants be scarce they readily turn their attention to corn, if any fields be near. Mating apparently occurs considerably in advance of egg-laying, though it also undoubtedly continues throughout the season of great- est activity, since pairs have been observed m coitu after the egg- laying season was well advanced. Oviposition was observed at Tappahannock chiefly during June, the earliest eggs being found on June 5. The beetles deposit their eggs a few inches below the surface of the ground wherever they happen to be feeding. It therefore appears that this insect spends practically its entire existence below ground. The beetles feed, mate, and oviposit, and the larvse complete their development below ground. Under ordinary summer conditions the eggs require from two to three weeks to hatch. When first hatched the larvee measure about 3 millimeters and when full grown about 32 millimeters or inches. The larvse require from six to eight weeks to reach maturity at Tap- pahannock, Va. Full-grown larvse were found from August 2 to October 2, but were most abundant the last week in August and the first week in September. The pupa stage lasts about two weeks under normal weather conditions. The first pupa found in the field at Tappahannock was August 16 in 1915 and August 12 in 1916. The latest field record was November 2, 1916. The majority of the old beetles die or disappear in midsummer; some stragglers, however, nearly always overlap the new generation. Such stragglers may be easily distinguished by their dull, opaque black, the new ones being highly polished. Adults of the new generation were found at Tappahannock as early as August 24, but the majority appear during the last half of Bui. 1267, U. S. Dept, of Agriculture Plate I The Rough-Headed Corn Stalk-Beetle A, Cornfield sit Tappahannock, Va., badly damaged by the rough-headed corn stalk-beetle (Eiutheola ruyicepi!) < photograph by W.J. Phillips); li, corn plant showing typical injury from adult of E. rngiceps (photograph by .1. II. Paine) Bui. 1267, U. S. Dept, of Agriculture Plate II The Rough-Headed Corn Stalk-Beetle A, Adult; B, egg when first deposited, inclosed within a ball of earth, the latter broken open to show egg within; C, egg after development has begun; D, end view of egg just before hatching (the dark area is the mandible of the larva showing through the eggshell); E, lateral view of egg just before hatching; F, larva shortly after hatching. (A, photographed by J. H. Paine; B-F, by W. J. Phillips) THE ROUGH-HEADED' CORl\T STALK-BEETLE 5 September. The new generation of beetles do not appear to be very active, remaining usually where they emerge, though they have been found soon after emergence feeding upon the culms of J uncus ejfusus and certain grasses of the genus P asp alum. EGG DESCRIPTION The egg when first laid is oblong, pure white, and perfectly smooth (PL II, B ). Subsequently the egg enlarges until it is nearly double its original size and changes its form until almost globular (PL II, B , (7, and E) . Eggs measured by the senior writer averaged about 2 millimeters in length and 1.5 millimeters in diameter. These were several days old and approximately full size. No meas- urements of freshly deposited eggs were made. METHODS OF COLLECTING AND INCUBATING A small number of the eggs used in the investigations were gath- ered in the field, but the greater number were obtained from the breeding cages. In the field the eggs were found in the ground, where it was usually possible to obtain them by digging. To obtain eggs which were known with certainty to belong to the present species, adult beetles were confined in suitable breeding cages, from which the eggs were gathered at regular intervals. Each cage consisted of a 12-inch standard-size flowerpot filled with finely-sifted soil and covered with a cylindrical wire-screen top ; the whole outfit had essentially the same form and arrangement as that portrayed by Davis (d, pi. 3, fig. 4) • The soil in these cages was kept mod- erately moist, and at« intervals, varying from a few days to a week and a half, was passed through a fine-mesh sieve. The meshes in this were fine enough to retain the eggs, which were then transferred to the incubating boxes. In the breeding cages the beetles were first fed by transplanting young corn plants to the cages, but as the labor of replacing the food plants every fewT days proved burdensome, a handful or so of corn kernels was buried in the soil. These proved to be a highly satisfactory substitute, the beetles feeding upon them as readily as upon the living plant. The receptacles used in incubating the eggs were rectangular tin boxes like those used by Davis for the same purpose (d, pi. 4, fig- ?) • These boxes were about three-fourths filled with finely-sifted earth, which was kept to the right degree of moisture by occasionally add- ing a few drops of water with a pipette. As the eggs were trans- ferred to a box each was placed in a small pit made with the blunt end of a pencil, and as the boxes were filled the lid was replaced and they were then kept in the shade. Usually the lid would so conserve the moisture originally in the soil that in most instances it was unnecessary to add more water during the period of incuba- tion. PLACE OF DEPOSITION The eggs are deposited in the ground, and apparently the females exercise no particular care in the choice of a place in which to 6 DEPARTMENT BULLETIN 1267, XJ. S. DEPT. OP AGRICULTURE leave their eggs. As the beetles themselves require a certain degree of moisture in their surroundings, they avoid very dry situations at all times. Naturally the eggs are most frequently encountered in places to which the beetles resort for the purpose* of feed- ing. For this reason they are most numerous in infested corn- fields, and in old pastures and grassy waste lands which constitute the normal habitat of the species. Field observations indicate that cornfields, especially if they happen to be well drained and are kept in a good state of cultivation, are generally very unfavorable situa- tions for the subsequent development of the larvae; for, although eggs are laid abundantly in com hills, well-grown larvae were rarely found in the same fields later in the season. This may be plausibly accounted for by the fact that the soil in well cultivated cornfields during periods of high temperature and drought is unsuitable for the development of the larvae. MANNER OF DEPOSITION The process of oviposition has not been observed. In most in- stances it appears that the eggs are deposited singly, although oc- casionally several may be found within a space an inch square. They are rarely inclosed in a clearly defined ball of earth. Possibly this may be due to the rather incoherent nature of the soils in which eggs were obtained in Virginia. At Charlottesville some experi- ments were conducted to ascertain whether the beetles were capable of forming such balls of ear^h by varying the moisture content of the soil and by adding clay to it. As a result a number of more or less firm balls were obtained, each inclosing a cavity containing a single egg (PI. II, A, C, D, E) , but the greater number of eggs were left loose in the soil, apparently with no attempt on the part of the beetles to inclose them in a ball of earth. All the earth balls obtained were found in soil that, had been fairly well saturated with water. This circumstance would indicate that the particles of which the earth balls are composed are held together only by the cohesive tenacity of the clay, and not by a glutinous secretion of the beetle. NUMBER DEPOSITED The number of eggs one female is capable of depositing under natural conditions is difficult to ascertain directly, but some experi- ments conducted at Charlottesville in 1915 provide data which with a certain degree of reservation may be used as the basis for an esti- mate. These data, indicate that the average deposition for each indi- vidual may vary from no eggs to rather more than three a day. Part of this variation may be accounted for by fluctuation of tem- perature. It has been repeatedly observed that high temperatures favor deposition, while low temperatures retard it. A part of the variation may also be attributed to the disturbance incidental to an examination for eggs. Usually the average rate of egg production for each individual varies, under particular summer conditions, from 1 egg in every 4 days to 2 eggs a day ; and it has been found that a similar range of variation in average daily production occurs if the figures are com- puted on the basis of a longer period, such as a month (or its equiv- THE ROUGH-HEADED CORN STALK-BEETLE 7 alent in clays), provided that no months later than September are taken into consideration. From these results it would appear to be a fair inference that on the average each female, under conditions similar to those existing in the experimental cages, is capable of depositing an egg every day during the normal breeding season. If the season lasts between one and two months, a beetle during this period ordinarily may deposit from 30 to 60 eggs. GROWTH As previously stated, the egg alter deposition enlarges until it is nearly or quite double its original size, and simultaneously changes its form until it is almost globular (PI. II, B-E ). Eggs measured at Charlottesville several days after deposition averaged about 2 millimeters in length and 1.5 millimeters in diameter. Unfortu- nately no measurements of freshly laid eggs were made, but the weights of eggs in different stages of growth were determined with the following results: In one lot of 15 eggs, all weighed within less than 48 hours after being deposited, a total weight of 0.02013 gram was obtained, an average of 0.001342 gram for each egg. Three days later this same lot weighed 0.04041 gram, an average of 0.002694 for each egg, practically twice the original weight. It was noted that one egg of this lot, on the second weighing, had not increased in size, possibly not having been fertilized, so that the average weight of an egg, at this time, was doubtless somewhat greater than the figures given indicate. In a second batch of 15 eggs, weighed when they were between 10 and 11 days old, the total weight obtained was 0.05052 gram, an average of 0.003368 gram for each egg, or approximately 21 times the average of an egg when deposited. In a third batch, consisting of 9 eggs estimated as 16 days old and nearly ready to hatch and 6 others which were at least 12 days old, the total weight obtained was 0.06538 gram, an average of 0.004359 gram for each egg, or approximately 3^ times the weight of a freshly deposited egg. No attempt was made to ascertain the cause of this increase in size and weight of the egg by determining its dry weight, but it is doubtless due to the absorption of water by the egg from the sur- rounding soil. The fact that the dead egg referred to above had not perceptibly increased in bulk after remaining in the soil for three days indicates that only the living eggs are capable of absorbing water. MOISTURE REQUIREMENTS It appears evident, as intimated in the preceding section, that the presence of a certain amount of available moisture in the soil is an essentia] prerequisite for development. The point was tested experimentally at Charlottesville, and it was learned that the eggs perish if kept in dry soil. It is to be regretted that no quantitative determinations of the moisture requirements of the eggs were made, as these would have been of value in explaining the conditions under which the species 8 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE occurs in' nature. The experience of the writers, however, both in the field and in experiments, indicates that an excess of moisture is almost as unfavorable for the eggs as a deficiency. The places in which the species normally occurs and in which it breeds most abundantly are so situated with reference to local drainage condi- tions that, although the ground retains a constant supply of avail- able moisture, it is never saturated with water for any considerable period of time during summer. It is doubtless these moisture re- quirements which underlie the marked tendency of the species to congregate in the lowlands bordering marshes and drainways and to be limited on the higher groiyid to local sunken areas in which the soil is rather slowly drained. INCUBATION Experiments on the duration of the period of incubation were conducted at Charlottesville. Owing to conditions attending the work, it was rarely possible to determine the precise time at which an egg had been deposited, and, for this reason, the results obtained are at best but approximations. On one occasion (July 26) a female was found in the act of depositing an egg. On August 9 this egg hatched, giving an incubation period of 14 days. There may be considerable variation in the time required for in- cubation. This may be accounted for by variations of temperatures, high temperatures accelerating and low ones retarding development. As a rule, the period of incubation under favorable midsummer con- ditions varies from two to three weeks. In the fall this time is greatly extended, extremes of from 35 to 50 days being reached in October and November. All eggs which had failed to hatch by the middle of November were buried in their containers in the ground and there kept over winter. In early April they were dug up and examined, but in all instances the eggs were dead. HATCHING The chitinized larval jaws may be seen through the translucent egg membrane (PI. II, D) several days before hatching takes place. When hatching occurs, the egg membrane appears to collapse and to split at a point close to the dorsal surface of the larva. In one instance the rupture of the membrane occurred in the vicinity of the thoracic region ; in another at about the level of the third abdominal segment. It evidently results from the contortions of the inclosed larva in an effort to free itself. After the membrane has split the larva continues its efforts, bending and extending its body at fre- quent intervals until it has finally managed to extricate itself, though occasionally portions of the membrane may adhere to the larva for a considerable time after hatching. In no instance did the larvae make any attempt to devour the egg membrane. The time consumed in the process of hatching was in one instance 7 minutes, in another 35 minutes. Plate II, F , shows the newly hatched larva and Figure 3 shows the relative size of the head and body immediately after hatching. THE BOUGH-HEADED CORN STALK-BEETLE LARVA DESCRIPTION The full-grown larva of Euetheola rugiceps (fig. 2.) is a robust, thick-bodied grub, with an approximate length of 32 millimeters (1£ inches) and an average thickness of about 6 millimeters. It is nearly pure white, deepening posteriorly to a dark gray or brownish tint, clue to the dark color of the vis- cera appearing throug h the trans- parent cuticle; the legs are yellowish amber ; the spiracles orange ; the head shield a distinctly reddish hue, closely approximating a bright shade of In- dian red. In alcoholic specimens these colors are invariably much obscured. The most distinctive morphological features of the larvae are found in the head shield and the last ventral seg- ment. The head shield (PI. Ill, A) is distinctly, even coarsely punctate,. piG. 2.— Fuii-grown larva of Euethe- the punctures being especially coarse r and dense on the portion immediately Walton) above the clypeus. The last ventral segment of the larva (PI. Ill, D) bears a some- what irregular, median, double row of modified bristles, each having the appearance of a denticle or minute spine. In the possession of this feature the larva of Euetheola i^agiceps is unique, so far as the writers are aware, among the Dynastini and agrees with the larvae of the genera Phyllophaga and Anomala, though in these the corre- sponding character is much more regular and clearly defined than in Euetheola (PI. Ill, D-G). In all other respects it resembles the type of larva normal to the Dynastini. METHODS OF COLLECTING AND REARING Owing to the pugnacious habits of the larvae it is best to place each in a separate receptacle when collecting, also to place a little vegetable mold or fine soil in the box to prevent the larva from rolling about and being injured. Attempts Avere made to rear the larvae in flowerpot cages, similar to those used to confine the beetles, but the results Avere disappointing. The failures Avere perhaps due, in part, to the fact that suitable food Avas not supplied the young larvae Avhen the cages Avere started, as the food requirements of the young larvae were then very imperfectly known; possibly, also, to inability to protect the larvae against cer- tain of their enemies. Ants frequently invaded the cages, and, as they are known to attack and kill the iarvae, were doubtless respon- .ibie in some measure for the unsatisfactory results obtained. The Fig. 3. — A'oung larva of Euetheola rugiceps immediately after hatching. Note rela- tive size of head and trunk, in comparison with Fig. 2. (Drawn by W. It. Walton) 940G10— 24- 10 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE larvae are often infested with certain mites, which are decidedly in- jurious to them and frequently cause their death. During the season of 1915 these mites were so abundant and widely distributed that it was found necessary to fumigate all samples of soil used in the breed- ing boxes. The salve-box method of rearing larvae recommended by Davis (5, p. 138 ) gave the best results and proved entirely satisfactory, once the peculiar needs of the young larvae were ascertained. The principal difficulties were to provide them with an adequate supply of suitable food and to protect them from mites and disease fungi. A few grains of wheat or corn were added to the boxes containing the young larvae, but the larvae made no attempt to feed upon either the grain itself or the plantlet issuing from it. Scarcely better results attended the use of the different kinds of manure, either fresh or old and thoroughly weathered. The fine, fibrous rootlets of the corn plant were also tried, but without success. Finally,, a satisfactory food was found in a well disintegrated, brownish plant residuum, or vegetable mold, a thin layer of which occurred fre- quently in an old pasture at Tappahannock, where the species bred abundantly. At first this material was gathered beneath tussocks of the common rush (J uncus effusus ) where it consisted of the broken and decayed culms of this plant, but subsequently was obtained with much less difficulty in connection with certain grasses belonging to the genus Paspalum and with Japan clover. It was the original intention to utilize this material, owing to its softness, merely as a medium in which the young and tender larvae might be kept with the least chance of suffering injury. At first a grain or two of wheat was added to each of the boxes containing this material, but it was soon noticed that the wheat was loft untouched, whereas the vegetable mold decreased rapidly in amount as the larva; grew and was re- placed by excrement. The wheat kernels were thereafter omitted and finely sifted vegetable mold alone used with entire success. A larva, after emerging from the egg, was carefully removed to a small tin salve box previously half filled with a quantity of the vege- table mold, finely sifted and slightly moistened. The young larva would invariably burrow into the mold and excavate for itself an irregular cavity, or cell, and there feed upon the surrounding ma- terial. When this had been consumed, the larva was temporarily removed from the box, the feces and other wastes cleaned out, and a fresh supply of mold added. As the larvae grew, the amount of mold consumed by them in- creased rapidly, necessitating frequent replenishing. The plan was adopted of substituting for the vegetable mold a kernel or two of corn, previously softened by soaking in water overnight. With larvae from half -grown to full-grown this proved to be a satisfactory substi- tute for the mold, and greatly lessened the work of caring for them. Fresh kernels were added only when the old ones had been almost consumed. Considerable difficulty was experienced in protecting the larvae from the minute mites previously referred to, specimens of which were identified by Nathan Banks as the hypopus stage of Rhizo- glyphus pkyUoxerae Riley. During the season of 1915 this pest was THE ROUGH-HEADED CORN STALK-BEETLE 11 extremely troublesome, and it required the utmost vigilance to pre- vent its gaining access to the breeding boxes. According to Mr. Banks, the mites are saprophytic upon decaying vegetable matter, but whatever may be their normal feeding habits, it is the uniform ex- perience of the writers, as well as of others who have worked with white grubs, that the presence of these mites in the breeding boxes is highly detrimental to the larvae. All soil or vegetable mold for use in breeding boxes was thoroughly fumigated with chloroform to kill all mites. Boxes infested with mites were emptied and sterilized in boiling water. To remove the mites from the larvae the latter were gently, but firmly, held between the thumb and index finger of the left hand and the mites loosened and brushed off under a binocular by means of forceps. Sometimes, to facilitate the removal of the mites, the larvae were plunged for an instant into a very weak solution of formaldehyde and then quickly washed in tap* water. This treat- ment appeared to cause the mites to adhere less tenaciously to their host, and also had a quieting effect upon the larva. For some reason — possibly the prevailing low temperatures of the season — these and other species of mites appeared to be unusually scarce in 1916, so that during that year these precautions were found unnecessary. FOOD AND FEEDING HABITS The experience of the writers both in the laboratory and in the field indicates that the normal food of the larvae consists chiefly of decayed and disintegrated vegetable matter. This vegetable mold does not usually occur as a distinct layer, being intimately inter- mixed with the surface soil ; but in the particular pasture of Tappa- hannock where most of the collecting was done the mold had accumu- lated as a practically pure layer on the surface wherever the plant cover was sufficiently dense to protect it from wind and from the trampling of stock. This would be particularly true of old pastures that had not been tilled for a number of years. Vegetable mold of the finest consistency usually occurred under the low, matlike growths of Japan clover ( Lespedeza stu'iata ), wherever these were dense enough to afford it adequate protection. In the layer of vegetable mold, or in the soil immediately underlying it, larvae of Euetheola rugiceps in all stages of growth were found in abundance, particu- larly where it was associated with clumps of Paspalum, a circum- stance that is doubtless connected with the fact that these grasses constitute the usual food of the adults. In most other localities where the larvae were found the layer of vegetable mold was not as extensive or as clearly defined as in the pasture at Tappahannock. The favorite haunts of the species ap- pear to be low or poorly drained areas where the plant growth ap- proximates that characteristic of the borders of marshes. In such areas the accumulation of vegetable detritus is relatively rapid. Attempts were made both at Tappahannock and at Charlottesville to rear the larvae upon cow manure in various stages of decay. Fresh manure appeared to be highly injurious to them, but old, dry, and well-cured manure, when slightly moistened, proved fairly accept- able, although the mortality among the larvae fed in this way was excessively high. It would seem probable that, while the vegetable constituents of manure may be suitable for the larvae, other portions 12 DEPARTMENT BULLETIN 1267, U. S. DEPT. OP AGRICULTURE may be toxic for them. This view is supported by results of field observations, which show that larvae of this species are only very exceptionally associated with manure. Thus in the old pasture at Tappahannock, where the larvae were abundant, they were never found beneath the droppings of cattle, although repeated search was made for them in such locations. Furthermore, they were no more frequent in fields that had been treated with manure than in those that had been left untreated. The junior writer has repeatedly searched for the larvae in fields to which manure had been added earlier in the season, but although the larvae of certain other scara- baeids, such as Ligyrus gibbosus (De G.), Dyscinetus trachypygus (Burm.), and Cotinis nitida (L-), were unusually common in such fields, those of Euetheola rugiceps were either entirely lacking or extremely scarce. Whether, under natural conditions, the larvae ever subsist upon living plant material is a question which can not as yet be answered. From the fact that the older larva in the breeding experiments were fed with kernels of corn, it would not be unreasonable to suppose that they may feed to some extent upon living plant material. The frequent association of the larva with grasses of the genus Paspalum suggests the possibility that they may feed upon the rootlets of these plants, though it is also possible that this association is purely acci- dental— a result of the parent beetles depositing their eggs in such spots while feeding upon the plants. Howard (7) and Titus (13) have inferred that the larva feed upon the dead and dying roots of the kinds of cultivated plants — sugar cane and corn— destroyed by the adult beetles. Titus, indeed, goes so far as to offer the suggestion that the object of the beetles in attacking sugar cane is less to secure food than to provide a supply of dead and decaying vegetation for the larva to feed upon. So far as corn is concerned, however, there can be little doubt that the beetles attack it primarily for food, and that if the destruction caused thereby is of benefit to the larva it must be a very indirect benefit. The junior writer has tested the capacity of the very young larva to feed upon dead and decaying corn rootlets, and, while the experi- ments were not sufficiently extensive to settle the matter fully, the results were entirely negative. It has been suggested that the larva may feed in decaying wood, as do those of some of the near allies of this species. Examination of old logs and stumps at Tappahannock for larva of Euetheola yielded only negative results, and it seems reasonably certain that they do not occur in such situations. In the experiments at Charlottesville an effort was made to rear the larva from forest leaf -mold, but, although they appeared to eat this, only a very small proportion of the larva tested lived beyond the earliest stages. There is no evidence that the larva ever feed upon such material under natural conditions. All attempts to find the species in timbered areas were unsuccessful. It is apparently limited to open situations. , GROWTH The larva on hatching from the egg is approximately. 3 milli- meters long; when fully grown the length is about 32 millimeters (1| inches). Growth is rapid, the larva attaining full size in from THE ROUGH-HEADED CORN STALK-BEETLE IB 5 to 8 weeks after hatching — usually this is about a week or 10 days before it is ready to enter the prepupa stage. DURATION OF THE LARVAL PERIOD The length of the larval period in the experimental series varied from 4 4 to 94 days. In the majority of cases, however, it falls be- tween 50 and 65 clays, a fair average being about 57 days. Instances in which the duration of this stage was greater than this mostly belong to those larvae which developed late in the season, when low temperatures retarded their growth. The earliest date at which larvae have been found at Tappahan- nock is June 19. This was in 1916, when a few were hatched from eggs collected in the field on June 5. In the breeding experiments of 1915, larvae were still being hatched as late as November, but this was evidently abnormal, as there is no evidence that any are ever hatched in the field later than the first part of August. The latest date on which the young larvae have been found in the field is August 12. This was in 1916, when two were obtained in the old pasture at Tappahannock. The latest date on which full-grown larvae have been observed in the field is November 2. This also was in 1916, when W. T. Emery recorded finding a few larvae of what he supposed to be this species in the same pasture. These specimens were unfortunately lost before their specific identity could be fully established. As the early fall of 1916 was unseasonably cold, it is not unlikely that such an extension of the larval period as is in- dicated by Mr. Emery’s observation may have occurred. Although it is possible that the latest developing larvae may, in some instances, fail to reach maturity before winter, there is no evidence that any ever survive until the following spring. Both field observations and experiments to test this possibility have given only negative evidence. At Charlottesville the larvae, each in its own box, were buried in a compost heap on the approach of winter, but when the boxes were dug out and examined in the spring all the larvae were dead. MOLTING Experiments to determine the number of molts and the duration of the periods between molts were made at Charlottesville. Each larva upon hatching was transferred to a salve box, the bottom of, which was covered with a disk of moist blotting paper on which were placed a few particles of old cow manure, which had been previously fumigated with chloroform. These experiments were begun August 14, 1915, and were continued throughout the fall and early winter. After September 25 the larvae were kept indoors, where they were subjected to artificial heat. * The mortality in these experiments was high, even after vegetable mold had been sub- stituted for the manure. Consequently only a very small propor- tion of the larvae completed their development. Owing to the late date at which these experiments were begun the time intervals recorded between successive molts can have little significance as regards the duration of these intervals under field conditions. 14 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE These experiments showed, however, that the larvse during their growth pass through two molts. There is, of course, a third molt at the close of the prepupa state. This result is in harmony with those obtained by a number of other workers in Coleoptera. Immediately after hatching, the head shield of the young larva is distinctly wider than the trunk (fig. 3) ; at this time it is quite soft and pure white. It hardens, however, within a day or two and assumes the normal red color. It then ceases to grow, but the trunk continues to expand and in time exceeds the head in thickness. Then when the larva hasi attained considerable size it undergoes its first molt. In the process the skin of the trunk splits lengthwise on the dorsal side, while the head shield becomes detached from it and is forced off the head in front. The new head shield is soft and pure white at first; it expands rapidly after the molt and in- a short time hardens, assumes the characteristic red color, and again exceeds the body in width. Succeeding molts are accomplished in the same manner. The larvse were frequently observed to devour the exuvium shortly after the molt had been completed. Table 1 gives data on molting experiments at Charlottesville, Va. Table 1. — Molting records of Euetlieola rugiceps, Charlottesville, Va., 1915 Serial No. Yl— . Y3-_. Y4... Y7__. Y8— . Y9._. Y10_ Y16_ Y17. Y62. Y63_ Y67. Y71. Y72. Y73. Y74_ Y75_ Y76. Y79. Larva hatched First molt j-Aug. 14 _ _ [Sept. 19 {Sept. 15 [Sept. 14 Sept. 20 [Sept. 16 {Sept. 18 (Sept. 21 /Sept. 16 \Sept. 21 JOct. 20 \ -do Aug. 27 (?) j-Sept.. 1 (?) j-Aug. 14 j-Sept. 25 jsept. 14 fOct. 19 lOct. 14 Oct. 11 lOct. 14 Nov. 3 /Oct. 18 \Oct. 21 Sept. 27- Second molt Third molt Nov. 4 Oct. 14 Oct. 12 Nov. 21 Nov. 30 Oct. 18 Oct. 20 Dec. 10 Nov. 13 __do Jan. 29 Nov. 19 Jan. 27 Emer- gence of adult Jan. 6 Jan. 18 Jan. 18 Mar. 9 The experience of the writers in Virginia indicates beyond much doubt that the normal habitat of all stages of Euetheola rugiceps consists of open grasslands on low or poorly drained areas of rela- tively heavy, dark-colored soils. The conditions prevailing in the habitat were most fully investigated in the vicinity of Tappahan- nock, but visits to other points in the State, from which the species has been recorded, show that essentially similar conditions charac- terize the habitat in all the localities examined. Similar conditions have also been reported by other writers. Thus Howard (7, p. 12) quotes a correspondent who wrote from Canton, Miss., that this insect was the worst corn pest on heavy, wet land he had ever experienced. Webster p. 159) states that in Tensas THE ROUGH-HEADED CORN STALK-BEETLE 15 Parish, La., and St. Francis County, Ark., corn on clay soils was damaged by the beetles. Sherman (il, p. J/4) records observations of the same character in North Carolina. In the unpublished records and correspondence of the bureau, ad- ditional observations to the same effect are recorded. One cor- respondent stated that at Dalton, Ga., corn attacked by the beetles was most severely injured in land that had been in meadow and pasture. Another correspondent, of Eutaw, Ala., reported them as doing considerable damage to corn in bottom lands. Some very interesting observations along the same line have been recorded in his field notes by George G. Ainslie, who investigated an outbreak of the beetles in western Tennessee and Kentucky. In a field at Savannah, Tenn., which he examined, Ainslie observed that the greatest damage to com was in the lowest part of the field. At Milan, in the same State, he notes that the beetles were most abun- dant in the lower and moister areas, while at Guthrie, Ky., he found that the greatest injury had been done in a field which adjoined a boggy spot overgrown with large sedges, rushes, and grasses. While the majority of observers agree in reporting the species as most numerous in heavy, moist soils, Comstock (3, p. 238), on the contrary, states that in the sugar-cane plantations of Louisiana the injury inflicted by the beetles is confined to those sections in which the soil is of a sandy, friable character, and is lacking in those where it is of a heavy, alluvial type. In the vicinity of Tappahannock, it was found breeding in a num- ber of more or less scattered stations, each of which was examined with regard to location, type of soil, and character of vegetation. These situations were, without exception, confined to the lower, nearly level lands which border the Rappahannock River and which represent a former flood plain. One of the breeding grounds most thoroughly studied at Tappa- hannock was the old pasture frequently mentioned in the foregoing pages, and known as “ Coghill’s pasture.” This pasture included about 20 acres, the greater part of which consisted of a rather heavy clay loam of a dark gray or slate color, and was about 2 miles back from the river. It had not been cultivated for at least 25 years and undoubtedly was of a marshy or swampy nature formerly, being considerably lower than surrounding cultivated fields. Many of the lower spots of this pasture had a thick cover of grasses under which was a thin layer of vegetable mold, where, as previously stated, many larvae of Euetheola rugiceps were found. The vegetation covering this tract was chiefly composed of species of grasses and sedges, the most abundant of which were those belong- ing to the following genera: Panicum, Paspalum, and Fimbristylis. There were heavy growths of Japan clover ( Lespedeza striata) in places; in the moistest spots were numerous tussocks of the tall rush (J uncos eff usus) . A second pasture about 2 miles southeast of Tappahannock re- sembled Coghill’s pasture in all essential respects. It joined a wood- land locally known as White Oak Swamp, and was really reclaimed swamp, so the junior writer was informed. Larva? were plentiful here, also, under the heavy growths of PaspaUim laeve. A third breeding ground was in a pasture close to the Rappahan- nock River, bordering a tidal marsh. This pasture was quite low, 16 DEPARTMENT BULLETIN 1267, U. S. DEPT. OE AGRICULTURE its highest point probably not exceeding 4 feet above tide, from which point it sloped gently toward the marsh. Most of the larvae of Euetheola rugiceps were found within a few yards of the marsh under growths of Paspalum laere, the soil at that point being moist but not soggy. The soil and vegetation here were essentially similar to those in the Coghill pasture. Larvae of Euetheola rugiceps were found in a number of other locations near Tappahannock and wherever found in numbers the locations were similar in all essential respects to the pastures pre- viously described. Besides occurring in what may be considered their normal habitat, larvae have been found in locations that are not entirely typical. Such occurrences seem very localized and are restricted to areas near the normal breeding grounds. Also, larvae seem rarely to reach maturity in well-cultivated fields. For example, in cornfields near old breeding grounds, and in which the corn was practically de- stroyed, very few full-grown larvae could be found. The junior writer carefully examined a hay field for larvae, the sod consisting chiefly of timothy, clover, and Bermuda grass. This field was near the Coghill pasture and had been in sod only three or four years. Across one end, in an area about 20 feet square, a large number of larvae of Euetheola rugiceps were found, while elsewhere in the field the larvae appeared very scarce. There seems Fm 4.— ventral view of Read re- £0 ]3e no satisfactory explanation of this gion of pupa of Euetheola rugt- . J r asps, showing structure of mouth singular occurrence at present. I he field fromphotograph by j^Paine) just mentioned was planted to corn the following year and there was consider- able injury from E. rugiceps , the greater part of the injury being in the vicinity of the spot where the larvae were so plentiful the pre- ceding year. This does not prove conclusively, however, that a good part of the injury was not due to migrating beetles from the old pasture. The soil in the timothy sod of the field just mentioned was a fine, rather sandy loam. Fine sandy loams appear to constitute the dominant types throughout most of the region bordering the Bappa- hannock River. These soils apparently harbor Euetheola. rugiceps only in the poorly drained areas that have become overgrown with wild grasses. In the opinion of the writers thorough cultivation combined with good drainage will eliminate E. rugiceps as a corn pest in such localities. PREPUPA In the beginning of the prepupa stage the larva ceases to feed and becomes relatively quiescent, the power of movement being retained only within the posterior half, which is capable of being bent forward beneath the thorax and then straightened out again. This movement may be repeated a number of times in rapid succession and is doubt- less of use in assisting the creature to enlarge the cavity or cell in which the pupa stage is passed, as well as in splitting the larval integument and thereby freeing the inclosed DUDa. During the pre- THE ROUGH- HEADED CORN STALK-BEETLE 17 pupa stage tlie larva lies on its back, in a slightly curved position, in the cavity formed by its movement in the soil. While still inclosed in the old integument, the larva undergoes its transformation into a pupa. This process is initiated by the withdrawal of the internal mass of the body from the larval integument at its hind end, which becomes greatly shriveled. Finally, when the pupal body has been formed, the larval skin splits along the dorsal line, revealing the fully formed pupa within. The latter frequently passes its entire exist- ence inclosed within the split larval skin. PUPA DESCRIPTION The pupa of Euethe- ola rugiceps measures on the average about 15 Fig. 5. — Ventral view of posterior end of abdomen of j male pupa of Euetheola rugiceps, showing sexual millimeters in length, characters. (Drawn by Henry Fox) and is pale buff. Its general form is shown in Plate IV, A, and certain of its structural features in Figures 4, 5, and 6, but probably its most distinctive pecul- iarities are those of the mouth parts (fig. 4). The mandibles are relatively stout, roughly triangular in outline, and with the apex forming a rounded angle. The labrum is quite wide transversely, and has its free edges regularly and evenly arcuate. The maxillary palpi are short, conical structures which are nearly vertical in position and have their tips projecting but slightly below the level occupied by the other mouth parts. Other distinctive characters are afforded by the shape of the post- coxal process of the prosternum, which is rather short, blunt, and constricted near the middle, and by the elytral pads, which are smooth, or, at most, only ob- scurely costate. Sexual characters in the pupa occur in the ventral surface of the last abdominal segment. In the male this bears a prominent hemi- spheric protuberance, the apex of which is slightly indented (fig. 5) ; in the female this structure is lacking, but instead there is a minute median projection of the anterior border of this segment into the segment in front (fig. 6). This process shows a pair of lateral, more fully chitinized areas which probably correspond to the genital plate of the adidt female. Fig. 6. — Ventral view of posterior end of abdomen of female pupa of Euetheola rugiceps, showing sexual characters. (Drawn by Henry Fox) 94051°— 24- O 18 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE DURATION OF THE PUPA STAGE The pupa stage may last from 9 to 44 days, but usually falls be- tween 10 and 19 days, so that two weeks may be said to be a fair average period for the duration of the stage under normal condi- tions. The precise length of this period unquestionably is deter- mined by the prevailing temperatures. Instances in which the length of the stage is much in excess of the normal pertain to individuals which have undergone, their development late in the season. The earliest date on which pupae have been obtained at Tappahan- nock is July 31, when there appeared in one of the breeding boxes a pupa which had developed from a larva collected at the same locality on June 30. The earliest record of actually finding pupae in the field is August 12. They appear to be most abundant during the last part of August and first half of September. Pupae have been found in the field as late as November and it seems quite probable that a small number may fail to mature before winter. There is no evidence that such pupae ever survive until the following spring, as all the pupae in the possession of the writers buried in the ground at Charlottes- ville perished during the winter. The adult of Euetheola rugiceps (PI. I, B ; II, A; figs. 7, 8) is a rather stout, jet black beetle, having an average length in Virginia and Tennessee material of from 13 to 16 millimeters. The surface in recently emerged individuals is highly polished,5 but is dull and opaque in old and worn ones. 5 Casey ( 1 , p. 187) in bis recent memoirs asserts that the body is “not very shining-,” and gives this as one of the characters distinguishing Euetheola rugiceps from another form from Honduras, which he describes as a new species, lionduranux. The writers are inclined to think from Casey's description that he had at hand only old individuals of rugiceps — doubtless collected during the spring, as the younger ones, collected in the fall,, are almost invariably rather highly polished and of an intense black color. Fig. 8. — Ventral view of adult Euetheola rugiceps, showing structural characters. (Drawn by Henry Fox) Fig. 7. — Euetheola rugiceps: Adult. (Drawn by Henry Fox) ADULT DESCRIPTION CORHEOTIOW SLIP apartment Bulletin 1267 , The Bough-Headed Corn Stalk-Beetle. 'LlS Cuts of ■^i6ure 7, page IS, and Bigure l4, page ^i^should ^tr^n0posed, as the former is that of Lfgyras gibto sus it ter is that of Euetheola rugioens. erie. 2?H0shoi ■m- • : •r bat • ■ THE ROUGH-HEADED CORN STALK-BEETLE 19 The head is rather short, its median length being about half its maximum width. Its surface is marked by numerous transverse, undulate rugulse (fig. 8) which are reduced to minute granulations on the front half of the clypeus and disappear on the occiput, which is quite smooth, except for a few sparse, shallow punctures. The clypeus has strongly oblique sides which are conspicu- ously margined and elevated. Immediately in front of the clypeal suture the head is crossed by a rather low, transverse ridge, or carina, which is broadly interrupted in the middle. The apical margin of the clypeus is almost trun- cate and rather short, being only about one-fourth the width of the base. It is dorsally reflexed and crested, the crest being interrupted in the middle by an oblique sinus, which separates the two conical processes — the so-called “ teeth ” — arising from the crest. These “ teeth ” in fresh specimens are rather high and sharp, but in old and worn specimens are frequently reduced to mere stumps. The mandibles are visible from above beyond the sides of the clypeus, and are very unequally bidentate, the anterior tooth, which is upturned at the apex, being much larger than the short, obtuse, posterior one. The pronotum is distinctly wider than long, and about twice as wide as the head. Its sides are broadly and evenly arcuate and narrowed slightly an- teriorly, the surface being smoothly and uniformly convex, bearing numerous coarse, annular punctures which are somewhat sparsely distributed through- out but rather more crowded on the sides than in the middle. The anterior and lateral borders are clearly margined, the posterior plain and feebly bi- sinuate. The antero-lateral angles are sharply produced anteriorly, while the postero-lateral ones are broadly rounded and obtuse-angulate. The scutellum, although rather small, is quite distinct, the surface being smooth, except for a few minute punctures. The elytra are but slightly longer than their combined width, which is not obviously greater than that of the pronotum. Each is longitudinally traversed by a number of slightly impressed, double rows of rather coarse, circular punctures, these giving the elytra a somewhat striate appearance (fig. 7). Anteriorly these punctures are frequently confluent and variolate. Outside of the double rows of punctures, the entire surface of the elytra is covered with numerous closely set and irregularly distributed punctures, which, for the most part, are essentially similar to those forming the double rows, but are reduced on the sides and apical half to minute, punctate impressions. The stridulating organs on the inner surface of the elytra are very feebly developed. The labium ( fig. 8) is considerably longer than wide, and is appreciably nar- rowed at its apical end, which is feebly bilobed and marked by sharply elevated lateral margins situated under the insertion of the palpi ; the basal half is rather strongly convex, without lateral margins, and bears on the sides nu- merous long stiff bristles, which are largely lacking toward the center. The prosternum (fig. 8) bears a stout, erect, cylindrical, postcoxal process or spine, the apex of which is almost flat and occupied by a smooth, padlike surface, the hind margin of which bears a conspicuous, radiating fringe of long, stiff bristles. The surface of the mesosternum and metasternum is nearly smooth, or at most but very sparsely and indistinctly pilose. The metasternum bears nu- merous shallow, circular punctures, larger and coarser on the side than in the middle, each of which is frequently provided with a minute, barely visible bristle or seta. The mesepisternum has its surface somewhat rugulose and bears a rather sparse covering of stiff hairs. The forelegs are relatively stout and are adapted for digging. The tibiae bear on the hind margin four distinct, toothlike projections, three being long, stout, and acute ; the fourth, or uppermost, much smaller and decidedly obtuse. SEXUAL CHARACTERS The last ventral segment of the male (fig. 9) bears near the anal margin a transverse fringe of short, stiff hairs which is broadly in- terrupted in the middle ; back of this interruption or hairless interval there is a short postanal fringe. In the female (fig. 10) the same character is also present, but there is no median break, the fringe being continuous. 20 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE A less obvious difference between the sexes is, as pointed out by Casey, in the form of the pygidium, which is slightly shorter, more convex, and more broadly rounded at the apex in the male than in the female. The male claspers (fig. 11) are symmetrical, each consisting of a vertical flange resembling that of Ligyrus gibbosus (fig. 12) but considerably smaller and slenderer and with the upturned process on its postero-lateral face more nearly basal, toothlike, and extending obliquely backward. The female genitalia consist of two pairs of almost flat plates — a large superior and a small in- ferior pair, the latter fringed apically with short hairs. A pubic process is lacking. TIME OF EMERGENCE The earliest date on which adults of the newly emerged generation have been ob- served under laboratory conditions at Tappahannock is August 13. This was in 1915, when an adult, reared from a larva collected June 30, appeared in one of the breeding boxes. The earliest date on which adults have been found in the field is August 24 (in 1915 and 1916). • The period of emergence extends throughout the last part of August and the whole of September and October, although ordi- narily relatively few appear to emerge later than the end of Sep- tember. The latest emergence of which there is record is of two individuals which developed in the breeding boxes early in November. In 1915 the majority of the beetles emerged between September 1 and Septem- ber 25, the period between September 10 and Septem- ber 20 being especially pro- lific in emergences. COLOR CHANGES When the adults emerge from the pupa they differ greatly in color from the typical mature beetles. The earliest changes in color take place during the closing days of pupal existence, while the adult is still inclosed within the pupal integument. These changes involve only the head and thorax, which at the time of emergence are already well chitinized and bright orange red. The elytra, however, at this time are quite soft and colorless, but assume a pale creamy hue within an hour or two, also becoming perceptibly firmer,* in a few hours more this changes to a bright orange. The following day, under normal conditions, the color of the elytra gradually grows darker, becoming a vermilion red or Indian red. In the meantime the head and thorax have been changing color Fig. 10. — Ventral view of tip of abdomen of adult female Euetheola rugic&ps, showing structural characters. (Drawn by Henry Fox) Fig. 0. — Ventral view of tip of abdomen of adult male Euetheola rugioetps, showing structural char- acters. (Drawn by Henry Fox) THE ROUGH-HEADED CORN STALK-BEETLE 21 rapidly and have become considerably darker than the elytra and are a deep purplish* red. The elytra soon acquire the same shade. The final stage naturally is the transformation of this color into the deep black of the typical beetle. Under favorable conditions these color changes are completed in from four to five days, but in cooler weather the time required to effect them may be greatly extended. Thus, in October and Novem- ber, beetles were frequently found to retain their red coloration for a period of two or three weeks. ACTIVITY IN THE FALL Pig. 11. — Lateral view of male clas- per of Euetheola rugiceps. (Drawn by Henry Fox) The adults appear to be much less active in the fall than in the spring. So far as the writers are aware, there are no records of the beetles having been taken at lights during this season. At Tap- pahannock, in the fall . of 1915, they were fre- quently observed on or immediately under the sur- face in the places where they had emerged. Al- most invariably they were to be found beneath clumps of their favorite food plants, Paspalum spp., boring into and cutting off the culms of these grasses. The junior writer never observed any of the beetles outside of their natural habitat at this time of the year, but W. T. Emery, who vis- ited the breeding grounds of the species at Tappahannock in early November of 1916, reported that he had seen a small number crawl- ing on an adjoining highway. Mr. Emery states that the day on which these beetles were observed was unusually warm and mild, a circumstance which doubtless accounts for their wandering abroad. HIBERNATION No systematic observations on the hiberna- tion of the beetles were made. So far as the available evidence goes, it indicates that hibernation takes place in the normal feed- ing ground of the species and in much the same manner as in other scarabaeids which pass the winter in the adult stage. On one occasion during the plowing of a timothy pasture at Tappahannock in February of 1916, the junior writer picked up a few beetles of this species. The depth at which they occurred could not have exceeded 8 inches and was probably less'. From the fact that some larvae reach maturity in cultivated fields, it is probable that many hibernate there, but they are insignificant in comparison with the much greater numbers that hibernate and emerge in the normal habitat of the species. Experience with beetles kept in cages outdoors, during the winter of 1915-16, indicates a heavy mortality among the hibernating beetles during this season in the latitude of Virginia. At both Charlottesville and Tappahannock only about a third of all the Fig. 12. — Lateral view of male clamper of Ligyrus gibboam. (Drawn by Henry Fox) 22 DEPARTMENT BULLETIN 1267, U. S. DEPT. OP AGRICULTURE beetles placed in the hibernating cages in the fall were living when the cages were examined in the spring. That an equally heavy mortality may obtain under natural conditions is indicated by the fact that where the beetles had been quite abundant in the fall of 1915, only a few could be found in the following spring. Doubtless if the species could be kept under constant observation. for a succes- sion of years, it would be found that winter conditions constitute one of the important factors controlling the destructive outbreaks of the species which seem to occur at rather long intervals. APPEARANCE IN SPRING The beetles usually begin to emerge from their hibernating quar- ters in the spring in late April or in early May, except in the most southern portions of its range. The precise time of emergence is governed by prevailing weather conditions. Comstock (3) states that they become active as early as the middle of March in Louisiana. At Tappahanock the earliest dates on which they have been seen abroad were April 23 in 1915 and May 1 in 1916. At Clarksville, Tenn., the junior writer first observed them at street lamps on April 18, 1917. McConnell recorded them (unpublished notes) as active and destructive at Greenwood, Miss., on April 23, 1913. Webster (lit) reported them damaging com in Louisiana on April 25, 1888. Becker informed the junior writer that complaints of injury by the beetles came in from southern Arkansas about May 1. MATING Mating of Euetheola rugiceps is practically coextensive with the period of its maximum activity. The earliest date on which the beetles were observed m coitu was May 13, 1915, at Sharps, Ya., while the latest date on which they were observed mating under natural conditions was June 20, 1915, at Tappahannock. In the case of those kept in breeding cages, mating was observed much later than this, one pair being observed m coitu as late as September 10. Mating normally takes place underground, though on one occa- sion a pair were found m coitu on the surface in a slight hollow at the base of a cornstalk ; they were also found mating in tin boxes in which they were placed during collection. It seemed to be im- material to the beetles whether soil was in the boxes or not. OVIPOSITION Oviposition was observed to occur at Tappahannock chiefly during June, the earliest eggs being found on June 5. It would seem probable that eggs may be deposited during July, but the writers have no records1 of obtaining any during that month. Most of the eggs are apparently deposited during the last half of June-. Beetles kept in cages under somewhat artificial conditions continued ovi- position, except for temporary interruptions due to the inclement weather, throughout the summer, and until as late as the last of September ; a small number of eggs even being deposited in October and in early November. In nature, however, such prolongation of the breeding season evidently does not occur, as field experience indicates beyond a reasonable doubt that practically all the beetles THE ROUGH-HEADED CORN STALK-BEETLE 23 of the egg-laying generation of the year have perished by the first of August. At Tappahannock, eggs, or recently hatched larvae, were found in hills of corn and in a layer of vegetable mold. In this vegetable mold they were often deposited at the base of tussocks of the common rush, beneath clumps of pasture grasses, especially those of the genus Paspalum, and under low mats of Japan clover. The eggs hatch within two or three weeks under normal summer conditions. ACTIVITY IN THE SPRING The beetles are unquestionably much more active and attract far more attention in the spring than during the fall. This is due doubt- less to the activities connected with feeding and reproduction. The beetles are rather sluggish, and if their needs are adequately met they apparently do not roam much. When, from any cause, their needs are not satisfied, they may come out of the ground and go elsewhere in search of more favorable locations, either by flight or by crawling away on the surface. Apparently the beetles fly only at night, when they are frequently attracted to lights, but the junior writer has repeatedly observed them crawling on the surface in bright daylight. From observations made on caged individuals, it would appear that the impulse to wander may come from lack of food as well as from the instinct to mate. Thus, in cages in which beetles were con- fined without food, they often came out on the surface, especially at night, and crawled up the sides of the cages, frequently attempting to take flight; whereas in adjoining cages, in which the inmates were plentifully supplied with food, it was a rare event for one to be found on the surface at any time. A beetle has occasionally been observed to emerge from a hill of corn in which all the plants had been killed and move off to another where the plants were intact. FOOD PLANTS AND CHARACTER OF INJURY BY THE BEETLE Euetheola rugiceps is best known as an enemy of corn and sugar cane, but there is reason to believe that these are not its normal food plants. During the fall of 1915 the junior writer found them feed- ing abundantly upon certain species of grasses belonging to the genus Paspalum. These grasses have since been found in every section visited by him in which the species has been found or from which it has been reported, and there is accordingly every reason to believe that they constitute the favorite food of the beetles. Beetles kept in confinement ate the plants eagerly. At Tappahannock the species of Paspalum fed upon were determined as P. laeve- and P. plenipilum. The large, coarse-stemmed forms, such as P. ftorida/num , do not ap- pear to be acceptable to them. The beetles attack these grasses in much the same manner as they do corn, forcing their way beneath the tufts, or coming up under them from below, and boring into the culms where the latter lie in contact with the ground. Sometimes the culms are cut completely off, but even when they are not entirely severed such a thin and broken bit of tissue is left connecting the parts that the portion beyond the injury quickly wilts and dies. In the fall of 1915 it was a common occurrence to find large patches of Paspalum which had been almost or quite completely destroyed by them. 24 DEPARTMENT BULLETIN 1267, U. S. DEPT. OE AGRICULTURE The beetles also feed upon the common rush {J uncus effusus). The culms of this plant form a dense tuft and are extremely tough and dry, except at the base, where they are somewhat tender. The beetles attack and cut them off at that point. Owing to the crowded condition of the culms at the base of the plant, it was not possible to detect the beetles at work, but they were found lying motionless in such situations and beside the broken and shredded culms. Tufts of the rush, from which all imperfect culms had been carefully removed, were transplanted to a cage containing the beetles and, when ex- amined several days later, were found to have a considerable number of their culms broken off and shredded in the same manner as those observed in the field. It would appear, however, that the beetles prefer the Paspalum grasses to the rush. Bermuda grass (Capriola dactylon ) also is probably eaten by the beetles, though very much less readily than Paspalum. This grass occurs practically everywhere throughout the entire coastal section of Virginia and is especially characteristic of the better cultivated areas. It abounds in many situations in which Paspalum is scarce or lacking. Indeed, it would seem that the chief danger of Euetheola Tugiceps perpetuating itself in farming districts and other places outside its typical habitat lies in the universal presence of this grass and the apparent ability of the pest to utilize it as food when no other is available. One would imagine that the hard and wiry stolons of Bermuda grass would scarcely prove very attractive; nevertheless the junior writer has repeatedly found them torn and frayed in the manner characteristic of injury by this species. Similar injury has also been caused by planting the stolons in a cage containing the beetles. Corn is attacked by the beetles only in the spring and early sum- mer when it is young. Later in the season the stalks become too hard for them to penetrate. The plants may be attacked as soon as they appear above ground, and are not safe from serious injury until they are fully waist high. The beetles are particularly fond of the apical growing point of the stalk, the so-called “ heart,” which is the most vital and important part of the plant. In the early stages of growth of the corn plant this structure forms a minute conical bud, situated below the surface of the ground in the center of the stalk. To reach this part the beetle bores; into the stalk at any point between the surface of the ground and tha point of attachment of the roots, making a large, ragged opening (PI. I, B). The work of the beetle is indicated above ground by wilting of the inner set of leaves, the outer ones retaining their rigidity for a considerable period after the other leaves have died. In a. somewhat later stage of growth, after the stalk proper has begun to elongate and has carried the terminal bud well above ground level, the injury done by the beetle boring into the stalk is usually less severe, only a more or less extensive part of the pith at this time being destroyed, the more vital growing part being out of reach of the beetles. At this time the stalk is also considerably thicker than before, and a beetle may finish feeding before it has destroyed enough of the vascular supply of the plant to interfere seriously with its functions. The chief danger to larger corn plants is naturally in the weaken- ing of the stalk, which may result in its being blown over or broken THE BOUGH-HEADED COEN STALK-BEETLE 25 off by strong winds. Plate I, A, represents a badly injured cornfield in the vicinity of Tappahannock. This field had been replanted several times. There were a number of fields showing such injury in the vicinity of Tappahannock. An interesting discovery in relation to the feeding habits of Eue- theola rugiceps is that it will feed readily on apples, either in breed- ing cages or in the field. This fact was first ascertained by Ezra Shackelford, at Tappahannock, who informed the junior writer that lie had found a beetle feeding on a fallen apple in the orchard. This observation subsequently was verified. All efforts to find the beetles feeding under natural conditions on common grasses other than Paspalwm spp., and Bermuda grass were futile, though in breeding cages they were induced to accept Panicum Undheimeri and Fimbristylis baldwmiana. The indications, how- ever, are that the beetles do not like these plants, and that they feed upon them only when deprived of all other food. Since ironweed ( Vernonia noveboracensis) was a common weed in the typical habitat of the species, an experiment was made to ascer- tain if the beetles were capable of utilizing it for food. The results were entirely negative. The junior writer found that the common ragweed {Ambrosia artemisiae, folia ) , the well-known food plant of Ligyrus gibbosus , was unacceptable to E. rugiceps in breeding cages, nor could they ever be found attacking these plants in the field. One or two correspondents of the Bureau of Entomology have stated that potatoes are occasionally injured by Euetheola rugiceps. To test this point, the junior writer on one occasion buried a few tubers in a cage containing a considerable number of beetles, but the latter apparently took no notice of them. At Tappahannock Euetheola rugiceps could never be found attack- ing the common grass locally known as goose-grass ( Eleusine indica ) , but to ascertain if the beetles are capable of subsisting on it a quantity was transplanted to one of the cages containing the beetles. "When examined a week later it was found that a number of the culms had been shredded to some extent at the base, but that in only a few was the injury serious. Evidently the beetles do not willingly feed upon this grass, but may possibly do so to a slight extent if unable to obtain more acceptable food. Since the beetles were found to be rather common in one of the timothy-clover pastures at Tappahannock, tests were made to ascer- tain if the beetles would feed upon these plants. In both instances they were untouched. It is probable, therefore, that beetles living in timothy and clover fields derive their sustenance from some other plant associated with them. This, in all likelihood, is Bermuda grass, which is usually common in such fields. • Occasionally the adults are accused of damaging rice. The writers have never had an opportunity to study the species in rice-growing sections and are unable to speak on this matter from personal obser- vation. Inasmuch as Euetheola rugiceps in one or two instances was reported as injuring rice, and subsequently proved to be the allied species Dyscinetus trachypygus , the writers can not avoid the sus- picion that all other reports of such injury may be cases of mistakes in identification. Superficially the adults of the two species are much alike and may be easily confused by persons not familiar with their distinctive characters. 26 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE DURATION OF THE ADULT STAGE The writers have stated that the adults of the overwintering gen- eration of Euetheola rugiceps perish, under natural conditions, by midsummer. In the experimental cages numerous beetles of this generation not only survived the summer but in some instances lived until late in the fall. This late survival possibly was due to the pro- tection which the beetles received from the extremes of heat and dry- ness to which in nature they are exposed. A heavy mortality among the beetles immediately after the mating season is evidenced not only by the large numbers of dead beetles found in the field at that time but also by the fact that corn planted late in the season — after June 1 — is almost invariably much less severely damaged than that planted earlier. The longevity of adults in the breeding cages was undoubt- edly due to the better care they received, as other adults, confined in similar cages but left exposed in the open to as nearly natural con- ditions as possible, perished in midsummer within a few weeks after the cages were started. This view is supported by the testimony of others who have had experience with this species, Howard (7) mentions a correspondent at Canton, Miss., who reported that previous to July 9 he had had little difficulty in finding the adults, but after a week of dry weather they had entirely disappeared. Sherman ( 11 ) also quotes a corre- spondent who, writing on June 14, reported that, although the beetles had been very numerous and destructive in his cornfields, he had noticed that within the last few days the dead beetles could be seen all about the field. He added that his corn crop had been so completely destroyed that the field was plowed up on June 1 and a new crop planted a week later but that this second crop remained uninjured. Apparently the duration of the adult stage in the latitude of Vir- ginia is from 9 to 11 months. The results obtained in experiments in- dicate that under exceptional conditions in nature the adult stage may conceivably last from a year to 14 months. SPECIES LIKELY TO BE MISTAKEN FOR EUETHEOLA RUGICEPS Euetheola rugiceps is often associated with other species of scara- baeid beetles which may be easily confused with it. Its most constant associates are its close allies, Ligyrus gibhosus (De Geer) and Dyscinetus trachypygus (Bunn.). For this reason these two species will be considered in somewhat greater detail than the remaining forms. LIGYRUS GIBBOSUS (De Geer) The life history of Ligyrus gibhosus is essentially the same as that of Euetheola rugiceps. The larvae of L. gibhosus develop more rapidly, consequently the adults of the new generation appear earlier in the fall than those of Euetheola rugiceps. The writers have never found L. gibhosus injuring corn in the field, nor could it be induced to feed upon corn in breeding cages. LARVA In general form, size, and coloration the larva of L. gibhosus _ re- sembles that of E. rugiceps. As in the latter the fully chitinized THE ROUGH-HEADED CORN STALK-BEETLE 27 head is of a distinctly reddish color but, unlike E. rugiceps , the head is smooth or at most but slightly rugulose, lacking almost entirely the deep punctures which are so conspicuous in the latter species. (PI. Ill, A and B .) Furthermore, in the larva of L. gibbosus there is no trace of a median double row of modified bristles on the last ventral segment, such as occurs in E. mgiceps. (PI. Ill, D and E.) PUPA The pupa of L. gibbosus (PI. IV, B ) is distinguished from that of E. rugi- ceps by certain characters associated with the mouth parts, by the form and position of the postcoxal process of the prosternum, and by the prominent bi- costate elytral pads (PL IV, A and B , and figs. 4 and 13). The mandibles (fig. 13) are much smaller and slen- derer than those of E. rugiceps , and are further characterized by the truncate, not angulate, apex which lies in contact with the nearly straight sides of the labrum. The maxil- lary palpi are also shorter and rather more rounded at the apex than in E. rugiceps. The postcoxal process of the prosternum is less nearly erect and the apex is rather more acuminate than in E. mgiceps. ADULT Ifi general form and size the adult of L. gibbosus (fig. 14) resembles that of E. rugiceps , but is usually distinguishable at a glance by its reddish brown color and by the distinctly hirsute character of its ventral surface. Occasionally adults are found in which the color is so dark as to be almost black. The most reliable differential character is the presence in L. gibbosus of a median pit or depression close to the anterior margin of the pro- notum, which is entirely lacking in E . rugiceps. In front of this pit is a blunt spine or tubercle. Other distinguishing characters of L. gibbosus are the absence of transverse rugulse and the presence of 1 < DiVw7ifWh y s i u'nvy *Fo x*a " " ' a continuous transverse ridge on the dor- sal surface of the head. The stridulating areas on the inner surface of the elytra are well developed in L. gibbosus and are capable of producing a low but audible sound, which is usually heard whenever the beetles are han- dled. In E . rugiceps the stridulating area is barely recognizable and is apparently functionless. Fig. 13. — Ventral view of head re- gion of pupa of Ligyrus gib- bosus, showing structure of mouth parts. (Drawn by Henry Fox from a photograph by J. H. Paine) 28 DEPARTMENT BULLETIN 1267, U. S. DEPT. OF AGRICULTURE DYSCINETUS TRACHYPYGUS (Burm.) The life history of Dyscinetus trachypygus agrees very closely with that of Euetheolci rugiceps. Development takes place at about the same rate in both species. Both may occur in similar situations, though D. trachypygus appears to be more tolerant of the products of organic putref action. Thus it has been taken in both adult and larval stages in compost heaps and in the vicinity of pig- pens, situations in which E. rugiceps has thus far never been found. There is no evidence that the adults of this species ever injure corn, as all experiments made to test this possibility yielded only nega- tive results. Farther south they attack rice, and for that reason the species has been given the popular name of “ rice beetle.” LAEVA that of E. rugiceps mainly m the entire absence of anything suggestive of a median double row of modified bristles on the last ventral segment (PI. Ill, F ). The surface of the front of the head is sculptured as in E. rugiceps (PI. Ill, A and C). PUPA The pupa of D. trachypygus (PI. IV, G ) is readily distin- guished from that of E. rugi- ceps by its longer and smoother head and by the form of the mouth parts. The mandibles are much longer and more slen- der than in either E . rugiceps or L. gibbosus , and terminate in a short, nearly truncate apex which lies in contact with the sides of the relatively small labrum (fig. 15). The maxil- lary palpi are unusually elon- gate, with an acute apex, and project considerably beyond the general level of the other mouth parts. The postcoxal process of the prosternum is more nearly oblique and rather more blunt than in E. rugiceps. ADULT D. trachypygus (fig. 16) may be readily recognized by its tooth- less mandibles and by the form and smoothness of the head. The latter is both longer and wider than in either Euetheola or Ligyrus Fig. 15. — Ventral view of head region of pupa of Dyscinetus trachypygus. (Drawn by Henry Fox) The larva differs from Bui. 1267, U. S. Dept, of Agriculture Plate 1 1 1 The Rough-Headed Corn Stalk-Beetle and Species Likely to be Mistaken for It A, Frontal view of head of larva of Euctheola rvgiceps, showing the strongly punctate surface; It, frontal view of head of larva of Ligyrus gibboSus, showing its relatively smooth surface; (', frontal view of head of larva of Dgscinelus trachypygus, showing its strongly punctate surface; 1), ventral view of anal region of larva of rugiceps, showing median double row of modified bristles on last segment; E, ventral view of anal region of larva of I). gibbosus (note absence of any trace of a median double row of modified bristles on last segment); A', ventral view of anal region of larva of !>. trachypygus (note entire absence of a median double row of modified bristle- and relative coarseness of bristles); G, ventral view of anal region of larva of Phyllo- phnga sp., showing sharply differentiated median double row of modified bristles on last ventral segment. (Photographs by J. II Paine) Bui. 1267, U. S. Dept, of Agriculture PLATE IV The Rough-Headed Corn Stalk-Beetle and Species Likely to b& Mistaken for It A, Pupa of Euetheola rugiceps; E, pupa of Ligyrus gibbosus; C, pupa of Dyscinetus trachy- pygus. (Photographs by J. II. Paine) THE ROUGH-HEADED CORN" STALK-BEETLE 29 and is characterized by its large, nearly rectangular clypeus which is separated from the epicranium by a distinct clypeal suture. The surface of the head is smooth except for a number of rather sparse punctures and is entirely devoid of either rugulee or a transverse ridge. CYCLOCEPHALA spp. The only stages in the life history of Cyclocephala likely to be confused with Euetheola rugiceps are the larva and the pupa. The larvae of Cyclocephala are somewhat smaller than those of E. rugi- ceps and of more yellowish hue. They also differ in the smooth and shiny surface and yellowish amber color of the head shield and in the entire absence of any trace of a median double row of modi- fied bristles on the last ventral segment. The distinctive character of the pupa was not fully studied in this investigation, but the possibility of confusing it with the same stage of Euetheola is largely eliminated by the fact that the two occur at entirely dif- ferent periods of the year. Thus, at Tappahannock, pupae of Cyclocephala were obtained from the last of May to the early part of July, whereas those of E. rugiceps were never taken before August. PHYLLOPHAGA spp. So far as adults of Phyllophaga are concerned there need be no difficulty in distinguishing them from Euetheola rugiceps , while the larvae may be recognized by the reduced size and triangular outline of the supraanal plate, the angular form of the anal slit, the smooth and shiny surface and yellowish color of the head shield, and the presence of a conspicuous, sharply defined, double row of modified bristles on the last ventral segment (PI. Ill, G). NATURAL ENEMIES The data on the predacious enemies of Euetheola rugiceps are very incomplete, and little of importance has been added in the course of this investigation. The underground habits of the species render it difficult to obtain direct evidence of predatory enemies, and only a very small proportion of the individuals collected were parasitized. At Tappahannock the fields infested with E. rugiceps were ob- served to be frequented by flocks of crows, grackles, and bobolinks, which were probably feeding upon the beetles, though direct proof of this was not obtainable. The Bureau of Biological Survey, however, has found specimens of E. rugiceps in the stomachs of the crow, meadowlark, and bluebird, and of species of the closely related genus Ligyrus in the stomachs of numerous birds. Among the possible insect foes of E. rugiceps may be mentioned several species of Carabidae (ground-beetles), Asilidae (robber- flies), and ants. Carabid beetles were normally common in under- ground situations and were of frequent occurrence in the places where E. rugiceps has been found, but the writers have no direct evidence that they attack or kill the latter, though it seems likely that the smaller larvae, at least, may at times be the victims of these pre- dacious beetles. As for the Asilidae, Titus (13) mentions the larva 80 DEPARTMENT BULLETIN 1267, U. S. DEPT. OP AGRICULTURE of a robber-fly ( Erax lateralis Macq.) as an enemy to this and other scarabaeids. At Tappahannock the junior writer observed the larva of a similar or closely related form preying upon larvee and pupae of a species of Phyllophaga, but never found it attacking those of E. rugiceps. Ants were found to attack and kill any larvae of the latter or of other scarabaeids that chanced to be exposed on the surface, and it is reasonable to suppose that they would do the same thing underground, an inference which is supported by the observa- tion that rarely, if ever, were scarabaeid larvae of any kind en- countered in the vh lity of ant colonies. In the junior writer’s experience the most frequent enemies are certain mites which attach themselves to the body surface. That these mites derive any nourishment from their host the writers are not prepared to assert. Nathan Banks, then of the Bureau of Ento- mology, to whom specimens of the mites were sent and who deter- mined them as the hypopus stage of Rhizoglyphus phylloxerae Riley, asserts that they are saprophytes, feeding upon decaying vegetable matter. Whatever may be the normal feeding habits of the mites, it is the experience of the writers as well as of other investigators (Davis, 5; Smyth, 12) that the presence of these and other mites is highly detrimental to the grubs and also to the pupae. Upon adults they appear to have little effect. The junior writer has observed adults almost literally encrusted with mites and apparently none the worse for the presence of their uninvited guests. Larvae and pupae are more susceptible, however, and it is the opinion of the writers that the high mortality in the larvae and pupae in the breeding cages and boxes was due in large measure to the mites. Larvae have been found in the field with the mites attached to them, so that it is not alone in the breeding boxes that they are attacked. At Tappahannock in the summer of 1915 these mites were very numerous and troublesome, but in the following year they had all but disappeared. Possibly such fluctuations in the numbers of the mites from year to year may be one of the factors in determining the rather sporadic and irregular manner in which destructive out- breaks of E. rugiceps appear to occur. The larvae and pupae were found occasionally to be infested with minute whitish nematode worms. Usually these were observed on the surface, where they tended to congregate in the intersegmental furrows, but sometimes an identical or closely similar type of nema- tode could be seen, through the transparent body wall, moving about in the body fluid. There are unquestionably two species of true parasites, one of which, a dexiid fly, W. R. Walton determined as Megapariopsis opaca Coq. The maggot of this fly feeds within the body of the larva until it is ready to form the puparium. Those reared by the writers bored their way out of the host shortly before changing to puparia. The other parasite was a hymenopterous insect, of which none was reared to the adult stage. For this reason the specific identity of the parasite was not determined, but it closely resembles Tiphia inomata Say, the best known probably of all the enemies of Phyllo- phaga as described by Davis (|, p. 15) and Smyth (12) . The young of this parasite is a thick white maggot, which during the time it is feeding lies in a transverse position on the dorsum of its host THE ROUGH-HEADED CORK STALK-BEETLE 31 immediately behind the head. The few specimens observed by the junior writer in tin salve boxes failed to give up adults. In the field, however, he has found on several occasions the cocoons of what he is inclined to think is the same form. These resemble in general the cocoons of Tiphia and, like the latter, are characterized by having the head shield of the host attached at one end. In a number of cases the head shield of E. rugiceps has been found attached to these cocoons, but adults were not reared from them. All stages of E. rugiceps , but more especially the larva and pupa, are subject to infection by a fungus, specimens of which were identi- fied by Dr. A. T. Speare, formerly of the Bureau of Entomology, as Metarhizium anisopliae. CONTROL MEASURES As has been shown, Euetheola~ rugiceps breeds mainly in low, moist, poorly drained areas that have been allowed to remain as waste or pasture lands for a considerable period of time. In fact under normal conditions these are apparently the only places where the pest breeds in sufficient numbers to constitute a menace to corn- fields. Land that is kept in a high state of cultivation, with fre- quent and systematic rotation of crops, furnishes an unfavorable breeding ground for this beetle. Very few beetles reach maturity in cultivated fields ; occasionally quite a number may be found breeding in temporary pastures or hay fields. The numbers of beetles develop- ing in such places, however, are insignificant compared with those breeding in the normal habitat of the species. ELIMINATION OF WASTE LANDS AND OLD PASTURES Knowing these facts, by far the most important means of control naturally suggests itself, namely, the elimination of all old waste and pasture lands. All such lands should be thoroughly drained and included in the regular system of rotation practiced for the re- mainder of the farm. If it seems most desirable to retain these lands for pasture, they should be broken up and reseeded every few years. This would be advisable if only as a matter of good farming, since in localities troubled with this pest pastures will become over- grown with weeds of many kinds in a few years at the expense of the more valuable grasses. The practices suggested will not only destroy the chief breeding, grounds of the pest, but will make these lowlands more productive and profitable. Such pasture lands when broken up should not be planted to corn the first year. As no other cultivated crop is injured by Euetheola rugiceps , some other crop can be substituted. The fol- lowing year corn may be planted, as there is but a single generation of the beetles a year. PASTURING WITH HOGS When old waste or pasture lands can not be drained conveniently and included in the rotation, the probability of injury resulting from the presence of these breeding grounds may be eliminated largely by pasturing hogs on such land every year, at least during August and September. The hogs will root out the grubs industriously. 32 DEPARTMENT BULLETIN 126 1, U. S. DEPT. OF AGRICULTURE EARLY PLANTING Since the depredations of the beetles appeared to occur mostly during May and June in 1914 and 1915, experiments were conducted in 1916 at Tappahannock to learn something of the possibilities of control by early planting. The earliest plantings were on April 7, and plantings continued at two-week intervals until June 19. Though the test was too short to be conclusive, the results indicated that May plantings suffered the greatest injury from Euetheola rugiceps. CHANGE OF ROTATION As previously stated, corn should not be planted after sod where there is the prospect of injury from the beetle. Besides the rough- headed corn stalk-beetle, sodworms and cutworms are always a source of danger to corn planted on old sod land. Therefore any system of rotation which obviates the necessity of following sod with corn helps to avoid several serious insect pests. FERTILIZERS The application of barnyard manure or commercial fertilizers is beneficial, because growth is hastened and the corn plants are thus enabled more quickly to reach a state where they are less likely to be injured seriously. HAND PICKING Hand picking is at best only a temporary expedient and in most cases very expensive. When a field of growing corn has become infested, however, there is no other hope of relief. Cheap labor sometimes may be employed to collect and destroy the beetles found in young corn. This work should be done principally when the corn is being either plowed or thinned. LATE SUMMER PLOWING The rough-headed com stalk-beetle enters the pupa stage during the latter part of August and it is in this stage that the insect is most easily destroyed, the least disturbance being sufficient to kill the pupae. For this reason, wherever possible, sod lands should be plowed the last week in August or the first week in September for Virginia but earlier than this for more southern localities. SUMMARY OF CONTROL MEASURES Eliminate* all old pastures or waste land, especially low, moist areas, and drain such lands thoroughly. Pasture hogs in waste or pasture lands that can not be conveniently drained and cropped. Plant corn early, say about April 2*0 for tidewater Virginia, and earlier for more southerly localities. Give liberal applications of barnyard manure or commercial fer- tilizers whenever practical. Employ cheap labor to collect and destroy the beetles when a field first shows injury. Do not allow corn to follow sod if possible to avoid it. Plow sod land in late summer and early fall in order to destroy the pupae of the rough-headed corn stalk-beetle. LITERATURE CITED (1) Casey, Thos. L. 1915. Memoirs on tlie Coleoptera, v. 6, 460 p. Lancaster, Pa. (2) Comstock, J. Henry. 1880. A destructive enemy to sugar-cane. In U. S. Comr. Agr. Ann. Rpt. for 1879. Report of the entomologists, p. 246-247. (3) . 1881. The sugar-cane beetle. In U. S. Comr. Agr. Rpt. for 1880, p. 236-240 ; also U. S. Dept. Agr. Special Report 35, 1881, p. 3-8. (4) Davis, John J. 1913. Common white grubs. U. S. Dept. Agr. Farmers’ Bui. 543, 20 p., 12 fig. (5) . 1915. Cages and methods of studying underground insects. In Journ. Econ. Ent., v. 8, no. 1, p. 135-139, pi. 3-5. (6) Garman, H. 1909. Corn pests. Ky. Agr. Exp. Sta. Bui. 145; pt. 2, p. 291-298. (7) Howard, L. O. 1888-9. The sugar-cane beetle injuring corn. In U. S. Dept. Agr., Div. Ent., Insect Life, v. 1, no. 1, p. 11-13, 1888. (8) LeConte, John L. 1856. Notice of three genera of Scarabaeidae found in the United States. In Proc. Acad. Nat. Sci. Phila., v. 8, p. 19-25. (9) Phillips, W. J., and Fox, Henry. 1917. The rough-headed corn stalk-beetle in the Southern States and its control. U. S. Dept. Agr. Farmers’ Bui. 875, 12 p., 8 fig. (10) Riley, C. V. 1880. New enemy to sugar cane. In Amer. Ent., v. 1, n. s., p. 130. (11) Sherman, Frankhn, jr. 1914. Insect enemies of corn. In N. C. Dept. Agr. Bui., v. 35, no. 5, whole no. 196, 56 p., fig. 14. (12) Smyth, Eugene G. 1917. The white grubs injuring sugar-cane in Porto Rico. 1. Life- cycles of the May-beetles or Melolonthids. In Porto Rico Dept. Agr. Journ., v. 1, no. 2, April, 1917, p. 47-92, pi. 2-9. Litera- ture cited, p. 88-90. (13) Titus, E. S. G. 1905. The sugar-cane beetle {Ingyrus rugiceps Lee.), [with notes on associated species]. In U. S. Dept. Agr., Bur. Ent., Bui. 54, p. 7-18, fig. 1-6. (14) Webster, F. M. 1890. Notes upon some insects affecting corn. In U. S. Dept. Agr., Div. Ent., Insect Life, v. 3, no. 7, p. 159-160. (15) Weed, Howard Evarts. 1895. Insects injurious to corn. Miss. Agr. Exp. Sta. Bui. 36, p. 147-159. (16) . 1895. A new corn insect. In Miss. Agr. Exp. Sta. 8th Ann. Rpt., p. 71-72. 33 ORGANIZATION OF THE UNITED STATES DEPARTMENT OF AGRICULTURE Secretary of Agriculture , Henry C. Wallace. Assistant Secretary Howard M. Gore. Director of Scientific Work E. D. Ball. Director of Regulatory Work Walter G. Campbell. Director of Extension Work C. W. Warburton. Solicitor R. W. Williams. Weather Bureau Charles F. Marvin, Chief. Bureau of Agricultural Economies Henry C. Taylor, Chief : Bureau of Animal Industry John R. Mohler, Chief. Bureau of Plant Industry William A. Taylor, Chief. Forest Service W. B. Greeley, Chief. Bureau of Chemistry C. A. Browne, Chief. Bureau of Soils Milton Whitney, Chief. Bureau of Entomology L. O. Howard, Chief. Bureau of Biological Survey E. W. Nelson, Chief. Bureau of Public Roads Thomas H. MacDonald, Chief. Bureau of Home Economics Louise Stanley, Chief. Bureau of Dairying C. W. Larson, Chief. Office of Experiment Stations E. W. Allen, Chief. Fixed Nitrogen Research Laboratory F. G. Cottrell, Director. Publications L. J. Haynes, in Charge. Library , Claribel R. Barnett, Librarian. Federal Horticultural Board C. L. Marlatt, Chairman. Insecticide and Fungicide Board J. K. Haywood, Chairman. Packers and Stockyards Administration- "{Chester Morrill, Assistant to the Grain Futures Administration J Secretary. This bulletin is a contribution from Bureau of Entomology L. O. Howard, Chief. Cereal and Forage Insect Investigations _ G. A. Dean, Entomologist in charge. 34 ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1268 Washington, D. C. October 16, 1924 RETURNS FROM BANDED BIRDS, 1920 TO 1923 By FREDERICK C. LINCOLN, Associate Biologist Division of Biological Investigations Bureau of Biological Survey CONTENTS Page Introduction 1 Organized Activities in Bird Banding 1 Regional Banding Associations 3 Returns Reported to the Biological Survey 4 Explanations of Tables S Tables of Returns 1 6 Index 54 WASHINGTON GOVERNMENT PRINTING OFFICE 1924 : vf' r Wyt \. ■ '?: * UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1268 Washington, D. C. T October 16, 1924 RETURNS FROM BANDED BIRDS, 1920 to 1923 By Frederick C. Lincoln, Associate Biologist, Division of Biological Investigations, Bureau of Biological Survey Page . 5 - 6 . 64 CONTENTS Page Introduction 1 Organized activities in bird banding 1 Regional banding associations 3 Returns reported to the Biological Survey 4 Explanation of tables. Tables of returns Index INTRODUCTION The marking of birds by means of numbered aluminum bands is resulting in the accumulation of information that will be valuable in solving many problems in ornithology. The experiment of Audubon, as long ago as 1803,* 1 in banding a brood of phoebes is now well known, while the interesting and valuable results obtained by Dr. Paul Bartsch, in 1902, in his work with black-crowned night herons in the District of Columbia,2 and by Dr. John B. Watson, in 1907, with sooty and noddy terns at Dry Tortugas, Fla.,3 have also been detailed in full. ORGANIZED ACTIVITIES IN BIRD BANDING For the credit of conceiving a broader scope for the banding method credit must be given in this country to P. A. Taverner and Dr. Leon J. Cole. Mr. Taverner, in 1903, distributed to interested persons a number of hand-made bands bearing the legend notify the auk ny, and a serial number. Comparatively few birds were thus banded, and only one return appears to have been recorded.4 Through the efforts of Doctor Cole, the New Haven Bird Club in 1908 issued a comparatively small series of bands, bearing the legend box z. yale sta. new haven, conn. ; in 1909, the legend was changed to read notify the auk new york. The bands used by the New 1 Audubon, John James. Ornithological biography, vol. 2, p. 126, 1834. 1 Bartsch, Paul. Notes on the herons of the District of Columbia: Smiths. Misc. Col., vol. 45, pub no. 141ft, quart. issue, vol. 1, pts. 1 and 2, pp. 104-111, pis. 32-38, lft04. * Watson, John B. The behavior of noddy and sooty terns: Pub. 103, Carnegie Inst. Washington, paper 7, pp. 187-225, pis. 1-11, March, 1909. * The Auk, vol. 23, p. 232, April, 1906. 94052°— 24f 1 2 BULLETIN 1268, U. S. DEPARTMENT OF AGRICULTURE Haven club were the first to be issued by any ornithological club in this country. Somewhat more than 5,000 of these bands were distributed during the first season after their adoption, and Doctor Cole, in commenting on the fact, expressed his satisfaction that “approximately one-fifth of these were used and are now [1909] being worn by a very considerable number of wild birds.”4 His paper on the subject was presented at the meeting of the American Ornitholo- gists’ Union, in New York City, in November, 1909, and emphasized the need of a permanent organization. The subject was again taken up at a dinner held at the Hotel Endicott on November 8, 1909, and the organization of the American Bird Banding Association was perfected at the meeting which followed. The legend on the bands was again changed, the new inscription reading notify a. m. museum, n. y. (sic), or (on the smaller sizes) notify a. m. n. h., n. y. Adequate funds for the purchase of bands and other supplies constituted a source of more or less diffi- culty, but through the activities of this association, and particularly through the untiring efforts of its secretary, Howard H. Cleaves, and the assistance of the Linnsean Society of New York (since April, 1911), the work was prosecuted until January, 1920. During this period an important paper was presented by S. Prentiss Baldwin which indicated the high character of the results that might be obtained from the systematic trapping and banding of birds.5 6 In January, 1920, the work of the American Bird Banding Associ- ation was formally taken over by the Biological Survey as an adjunct to its investigations. The importance of this method of obtaining information relative to the migrations and life histories of birds is being fully demonstrated. Plans were at once formulated by the Biological Survey for the intensive development of the project, but putting them into full effect was delayed until an adequate supply of bands could be assured. A manufacturer was finally found after a long search, who was sufficiently interested to install and perfect the necessary ma- chinery. With a supply of bands available to make a thorough study, the bird-banding project has been developed steadily through an increasing number of cooperators, whose work is done without compensation. The work now bids fair to become one of the most effective means for gathering precise information relating to bird migration and to the life histories of many species. Studies conducted with many individuals of a species at different trapping stations — thereby check- ing the work of the different operators— are providing a wealth of indisputable information which will in some cases no doubt completely revolutionize some accepted beliefs. It is generally conceded that these results could be obtained in no way other than by the banding method as furthered by the establishment of numerous trapping stations. 4 Cole, Leon J. The tagging of wild birds; report of progress in 1909: The Auk, vol. 27, no. 2, p. 157, April, 1910. 6 Baldwin, S. Prentiss. Bird banding by means of systematic trapping: Abstr. Proc. Linnaean Soc. New York, no. 31, pp. 27-56, pis. 1-7, 1919. RETURN'S FROM BAUDED BIRDS, 1920 TO 1923 3 REGIONAL BANDING ASSOCIATIONS Because of the great difficulty of coordinating the activities of large numbers of cooperators in various parts of the United States and Canada, the expediency of grouping them into regional associations, each with its proper officers, was favorably considered. The plan at first contemplated was the allotment of territory along natural geographic lines as far as possible, although, unfortunately, it has been necessary to make the boundaries conform to political lines, because of the State permits that must be obtained. The system has been satisfactory, however, and under the guidance of experienced ornithologists the development of the local associations has been rapid. Northeastern Bird Banding Association. — First of these local asso- ciations was the present Northeastern Bird Banding Association, originally organized as the New England Bird Banding Association, on January 17, 1922, the result of the interest and tireless energy of Laurence B. Fletcher, of Boston, Mass., who has since served in the capacity of secretary. E. H. Forbush, State ornithologist of Massa- chusetts, was elected the first president. The territory covered includes the New England States, the Maritime Provinces of Canada, and Quebec. Inland Bird Banding Association. — The second regional organiza- tion to be launched was the Inland Bird Banding Association, formed on October 24, 1922, at the Chicago, 111., meeting of the American Orni- thologists’ Union. S. Prentiss Baldwin, whose work in systematic trapping opened a new field of bird banding, has served continuously as president, and William I. Lyon, of Waukegan, 111., a pioneer in the work, has served as secretary. This association is coordinating the work in the territory tributary to the Mississippi River, extend- ing from the Allegheny Mountains to the Rocky Mountain States and British Columbia. Banding Chapter , Cooper Ornithological Club. — The Cooper Orni- thological Club next took an active interest in the subject. Because of the fact that the existing organization was devoted primarily to the study of western birds, it was not considered advisable to promote a separate association. A “ Banding Chapter of the Cooper Club” was accordingly formed, with J. Eugene Law, of Altadena, Calif., as chairman. The territory covered by the chapter includes the entire Pacific coast area, including Alaska, together with the Rocky Moun- tain States. Eastern Bird Banding Association. — With three organizations func- tioning, there remained unallotted only the territory on the Atlantic coast south of and including New York, and the Province of Ontario. It was accordingly considered important to form an association of the cooperators m this region. At a special meeting of the Linnsean Society of New York, held at the American Museum of Natural History on April 24, 1923, the subject was fully discussed and Mr. Baldwin was authorized and requested to take care of the preliminary arrangements. As a result, the Eastern Bird Banding Association has been organized, with Dr. Arthur A. Allen, of Cornell University, Ithaca, N. Y., as president, Mrs. J. E. Webster, of East Orange, N. J., as secretary, and Rudyerd Boulton, of the University of Pittsburgh, as executive secretary. 4 BULLETIN 1268, U. S. DEPARTMENT OE AGRICULTURE So great is the extent of the territory assigned to most of these regional associations that further subdivision will ultimately be necessary. This is being anticipated and will be accomplished when the cooperators in any geographic area become sufficiently numerous and leaders for new groups can be developed. There are already many station operators of experience who will, no doubt, be capable of furthering such activities. RETURNS REPORTED TO THE BIOLOGICAL SURVEY Facts relative to the 1,746 returns that were received by the Bio- logical Survey during the period from January 1, 1920, to June 30, 1923, are set forth in this report in tabular form, in order that coop- Fig. 1. — Localities from which bands have been returned that were attached to mallard ducks at the Sanganois Club, Browning, HI., during the period February 28 to March 25, 1922, and September 27 to December 15, 1922. In the Mississippi Valley one spot on the map may indicate several returns. A study of the migration of this important game bird is receiving special treatment by means of the banding method and, as illustrated, information already received shows a dispersal from the point of banding over a vast region that extends from the coast of Texas and Louisiana north to central Manitoba and Sas- katchewan, with an occasional mallard traveling east to the Atlantic coast erators and others interested may the more readily use the available information for present purposes and as a basis for further investi- gations. In addition, it is believed that the meager results obtained from the banding of some species will serve as a stimulus for those who may be sufficiently ingenious to devise new and effective con- trivances for the capture of birds of these species in larger numbers. The tables in this report are presented without discussion, for the reason that it is deemed inexpedient to attempt detailed comment for all groups when it is obvious that the material will permit a wide range of interpretation. This will depend not only upon the charac- ter of the returns obtained for different species, but also upon ecologi- cal, meteorological, and other factors that demand full consideration. Bui. 1268, U. S. Dept, of Agriculture Plate I B24054 Fig. I. — Spring-pole Trap Duck station at Cuivre Island, Mo., operated by L. V. Walton B24074 Fig. 2. — Automatic Pen-trap Duck station maintained at the Sanganois Club, Browning, 111., by the writer. The trap contained 93 mallards, pintails, and black ducks when this picture was taken Bui. 1268, U. S. Dept, of Agriculture Plate II B2437 Fig. I. — Collapsible Drop-trap Used under cover for protection from the weather, at the station of Richard B. Harding, Cohasset, Mass. B2438 Fig. 2. — Collapsible Drop-trap (Close View) The trap illustrated above, with cover removed to show food hopper, by means of which a constant supply of bait is maintained during days when the trap is not operated. At such times birds are free to enter and leave at will Bui. 1268, U. S. Dept, of Agriculture Plate III B2439 Fig. I. — Tree-trunk Trap Operated at station of Richard B. Harding, Cohasset, Mass. B2367 Fig. 2. — Folding Warbler-trap Used by T. Donald Carter and R. TI. Howland in conducting studios with Brewster warblers. The trap contains an adult male of this species Bui. 1268, U. S. Dept, of Agriculture PLATE IV B2436M Fig. I. — Sliding-door Trap Operated at the station of B. S. Bowdish, Demarest, N. J.; 119 purple finches were taken in this trap on a single day B32435M Fig. 2. — Window Feeding-station and Trap A protected window trapping station operated by R. E. Horsey, at Highland Park, Rochester, N. Y. 5 RETURNS FROM BANDED BIRDS, 1920 TO 1923 ^ For instance, what might prove an excellent method of treatment for returns from the mallard duck (Fig. 1) would not be applicable in the case of the song sparrow. With a sufficient accumulation of information from different parts of the country, it will be possible eventually to prepare reports which will constitute distinct contributions to our knowledge of the species involved.* * * 6 The number of recruits to this new method of ornithological research is constantly increasing, and it is gratifying to note that many cooperators are directing their activities toward the solution of definite problems concerning certain species. EXPLANATION OF TABLES The various species are presented in the order of the “A. O. U. Check-List.” 7 The records under individual species are arranged numerically under the names of the operators, grouped alphabetically by States, which also are in alphabetical order. Thus under “ Tree sparrow,” the table begins with returns from birds banded in Connect- icut by cooperators listed alphabetically. This is followed by Illinois and Massachusetts, with the returns from the cooperators in the respective States, similarly arranged. This method of treat- ment has been adopted because any studies based on this material should place initial emphasis upon the locality (State or Province) where the birds were originally banded. Numbers preceded by an asterisk (*) are those of the American Bird Banding Association, all others being those of the Biological Survey. These records have been obtained mostly from birds marked since January, 1920, as the remnant of the association bands have been issued by the Biological Survey. In a few cases, however, it has appeared advisable to give the complete history of a bird that has yielded a series of returns of special interest, and in these instances return records received by the association prior to 1920 also are included. Numbers preceded by a dagger (f) indicate recently banded birds found dead at or near the place of banding. Unusually late dates in these “ short-time returns” are often due to the fact that the carrier was not found until some time after its death.8 Following is a list of known causes of such deaths : Due to — Cats. Shrikes. Red squirrels. Ground squirrels. Weasels. Rats. Owls (species ?). Sparrow hawks. Cooper hawks. “Grass” snakes. Gopher snakes. Black snakes. “Hawk or gull.” Due to — Blue jays. Hunters. Capture in — Herring net. Fish trap. Muskrat trap. Saturation of plumage with fuel oil. Crowding of fledglings out of nest by cowbirds. Inclement weather. Desertion of fledglings by parents. Accidents due to traps or operators. e One preliminary report of the migrations of certain ducks, as deduced from handing returns, has been prepared by the writer (Trapping ducks for banding purposes: The Auk, vol. 39, pp. 322-334, pis. 11-14, map, 1922;, and he has in progress more elaborate communications dealing with these birds, together with the herring gull and Caspian tern. 7 Check-List of North American Birds of the American Ornithologists’ Union, third edition (revised), 1610. * Martin No. 104232 (p. 4.r») was found dead when the colony house was cleaned out — long after the parents had left. 6 BULLETIN 1268, U. S. DEPARTMENT OE AGRICULTURE Of death-dealing agencies, by far the most numerous are cats and shrikes, with squirrels of various species next. Accidents at trap- ping stations that may be attributed to carelessness, inexperience, or improperly constructed traps have been remarkably few. Natur- ally some accidents would occur where inexperienced persons began handling large numbers of small birds, caught in traps which in some cases were either purely experimental or crudely constructed. The fact that such mishaps have been so few offers abundant justification of the bird-banding method and shows the care and gentleness exercised by cooperators. Information as to the cause of the death of banded birds when forwarded to the Biological Survey yields interesting light on the subject of bird mortality and its causes and is thus available for study purposes. Symbols and abbreviations used in these tables, in addition to the asterisk (*) and the dagger (|), already explained, are as follows: , male bird; $ , female; juv., young (juvenile) ; im., immature. TABLES OF RETURNS Loon: Gavia immer No. Age and sex Banding record Return record Operator Locality Date Date Locality 201486 H. A. MeGraw Altoona, Pa Apr. 30,1923 May 27, 1923 Balmy Beach, Simeoe County, Ontario. Glaucous-winged Gull: Larus glaucescens 200S90 juv_ Theed Pearse. Gulf of Georgia, July 30,1922 Feb. 16, 1923 Pacific coast, 86 mi. N.of British Co- Vancouver, British Co- lumbia. lumbia. 200994 jllV- Dec. 15, 1922 Redonda Bay, British Columbia. Herring Gull: Larus argentatus 100633 juv_ Ernest Joy Grand Manan, Aug. 16,1921 Jan. 4, 1922 Jamaica Bay, N. Y. New Bruns- wick. 100646 juv_ do do Oct. 18,1921 Maceo Bay, Charlotte County, New Bruns- wick. 100698 do Aug. 18,1921 Apr. 9, 1922 San Antonio Bay, Tex. 100700 do do June 18, 1922 Rockaway Point, N. Y. *25386 juv_ Wm. Pepper.. Little Duck Is- July 19,1915 Aug. 1920 Cape Porpoise, Conn land, Me. 202213 juv_ W. S. McCrea. St. James, Mich. July 16,1922 Jan. 18,1923 Wickliffe, Ey. 202237 do do__ do Nov. 30, 1922 Ennis, Tex. 202248 juv_ do A do__ do Jan. 12,1923 Brunswick, Ga. RETURNS FROM BANDED BIRDS, 1920 TO 1923 Bing-billed Gull: Larus delawarensis 7 No. Age and sex Banding record Return record Operator Locality Date Date Locality 5553 juv. Francis Harp- er. Lake Atha- baska, Al- berta. June 28, 1920 Oct. 2, 1920 Lesser Slave Lake, Al- berta. Caspian Tern: Sterna caspia imperator 202257 juv. W. S. McCrea. St. James, Mich. July 28,1922 Sept. 24, 1922 Traverse City, Mich. Common Tern: Sterna Mr undo *1258 juv. J. C. Phillips. Eastern Egg July 3, 1913 Aug. 1917 1 Niger River Delta, South Rock, Me. Nigeria, West Africa. *54886 F. P. Cook Brigantine Is- July 27, 1920 May 20, 1922 Paria Bay, Blanchisse- land, N. J. use, Trinidad. 104872 juv. Edwin Beau- Bay of Quinte, July 14, 1922 Aug. 15,1922 Rochester, N. Y. (12 mi. pre. Ontario. W.). Gannet: Moris bassana 207105 juv. H. L. Stod- Bonaventure Is- July 31,1922 May 2,1923 Alberton, Prince Ed- dard. land, Quebec. ward Island. 207236 Sept. 15, 1922 207269 Nov. 27, 1922 tia. White Pelican: Pelecanus erythrorhynchos 100553 juv. A. F. Walther. Morse, Sas- Oct. 22,1921 Oct. 27,1921 Lindsay, S. Dak. katchewan. 1201809 juv. H. B. Ward... July 26,1922 Yellowstone Lake, Wyo. Lake, Wyo. 201815 do Oct. 9, 1922 201820 do Sept. 1922 1201826 do t201831 do Do. 201835 juv. do Oct. 5, 1922 1201841 juv. do 201843 juv. do do Feb. 26, 1923 201850 juv. do Oct. 1, 1922 mi. NW.). 201860 do Oct. 11,1922 201873 juv. Oct. 10,1922 Aurora, Utah. 1 Reported under date of Nov. J8, 1920. 8 BULLETIN 1268, U. S. DEPARTMENT OF AGRICULTURE Mallard: Anas platyrhyncha No. Age and sex Banding record Return record Operator Locality Date Date Locality tl01043 fl01045 102401 102402 9 do do do Dec. 13, 1922 Beardstown, 111. 9 do do do Nov. 28, 1922 katchewan. Browning, 111. 9 do do do do Do. c f do do__ do Dec. 20,1922 BlufE City, 111. cf do do do Dec. 10, 1922 Crane Lake, 111. c? do do do. Dec. 7, 1922 Bath, 111. d1 do do do May 3, 1923 s Oxford House, Manitoba. 9 do do do Dec. 30,1922 Bath, 111. (4 mi. S.). 9 do do do Dec. 7, 1922 Beardstown, 111. cf do do Nov. 30, 1922 Dec. 9, 1922 Beardstown, 111. (10 mi. 9 do do do Dec. 25,1922 NE.). Snicarte, 111. 9 do do do Jan. 10,1923 Cash, Ark. (6J mi. S.). 9 do do do Dec. 19,1922 Beardstown, 111. (6 mi. 9 do do do Dec. 27,1922 NE.). Astoria, 111. d1 do do Dec. 2, 1922 Dec. 16,1922 Beardstown, 111. o’ do do do Dec. 26,1922 Havana, 111 . (5 mi. N E .) . d1 do do do. Dec. 18,1922 Browning, 111. c? do do do Jan. 20, 1923 Lauderdale County, Ala. Beardstown, 111. c? do do do Dec. 16,1922 9 do do do Dec. 22,1922 Crane Lake, 111. 9 do do Dec. 11,1922 Dec. 30,1922 Browning, 111. 1£ mi. SE.). Frederick, 111. d" do do Dec. 3, 1922 c? do do do do Beardstown, 111. (7 mi. 9 do Dec. 29,1922 N.). Bath, 111. (9 mi. S.). Browning, 111. (lJmi.S.), Bradley County, Ark. cf do Dec. 4, 1922 c? do 9 9 do do do Dec. 13, 1922 Beardstown, 111. 9 do Dec. 20,1922 Dec. 25,1922 Jan. 4, 1923 Cass County, 111. Bath, 111. 9 do 9 do do Bellvllle, Tex. approximate. 14 BULLETIN 1268, U. S. DEPARTMENT OF AGRICULTURE Mallafrd — Continued Age and Banding record Return record sex Operator Locality Date Date Locality 231273 9 F. C. Lincoln. Browning, 111 Dee. 4, 1922 Dec. 25,1922 Snicarte, HI. (J mi. N.). 231278 cf do. do Dec. 5, 1922 Dec. 12,1922 Crane Lake, 111. 231281 cf do__ do do Dec. 10,1922 Snicarte, 111. 231283 cf do do do.. Jan. 27,1923 Tilton, Ark. t231288 cf do do do 231304 9 do do __ do... Dec. 25,1922 Ark. (Cache R.). 231306 9 do do do Dec. 30,1922 231309 9 do do do Dec. 8, 1922 Beardstown, 111. (3 mi. NE.). 231310 9 do Dec. 13,1922 231314 cf do do Dec. 6, 1922 Dee. 15,1922 White River, Ark. 231315 cf do Dec. 17,1922 231321 cf do do do... Dec. 13,1922 Snicarte, 111. 231323 cf do do Dec. 14,1922 231325 9 do Dec. 28,1922 231326 9 do do do Dec. 20,1922 Bath, HI. 101682 cf E. A. Mcll- Avery Id., La.. Feb. 9, 1922 Aug. 17,1922 Cumberland Lake, Sas- henny. katchewan. 101684 cf Feb. 15,1922 Dec. 16,1922 101690 cf Feb. 9, 1922 Nov. 21, 1922 101695 cf do do do Jan. 9, 1923 Do. 203305 9 JohnBroeker4. Portage des Jan. 22,1923 Apr. 29,1923 The Pas, Manitoba (60 Sioux, Mo. mi. E.). 203332 cf do do Jan. 24, 1923 May 1,1923 The Pas, Manitoba (8 mi. E.). t203348 9 do Jan. 27,1923 Mar. 6, 1923 203365 Jan. 31,1923 203385 cf do f203409 9 do Feb. 3, 1923 Mar. 6, 1923 203420 cf do do ... .do Feb. 18,1923 Gale, HI. 203548 9 do do Mar. 4,1923 May 10,1923 Reindeer R., Saskatche- wan. 203683 9 Mar. 24, 1923 May 16,1923 101302 L. V. WaltonA Cuivre Id.,Mo__ Jan. 7, 1922 Dec. 22,1922 Holcomb, Miss. 101305 do Nov. 16, 1922 101313 Jan. 7, 1922 Nov. 24, 1922 101319 Jan. 11,1922 101320 9 Nov. 21, 1921 Dec. 18,1921 101327 do do Oct. 12,19222 3 4 wan (15 mi. NW.). 101328 Jan. 27,1922 101329 do do Dec. 26,1922 S.W.). 101355 do Jan. 7, 1922 Jan. 18,1923 E.). 101365 do do _ Jan. 11,1922 Sept. 30, 1922 Portage la Prairie, Mani- toba (13 mi. N.). 101374 do do... Jan. 8, 1922 Dec. 7, 1922 Clarendon, Ark? (7 mi. 101376 do do Jan. 7, 1922' Nov. 12, 1922 S.). Lake Williams, N. Dak. 101377 do do Jan. 8, 1922 Feb. 8, 1922 Morse Mill, Mo. 101388 do do __ Jan. 12,1922 Sept. 16, 1922 Jamesville, S. Dak. 101391 do Jan. 8, 1922 Nov. 27, 1922 Chesterfield, Mo. 101395 do do Jan. 7, 1922 Oct. 15, 1922 Ellendale, N. Dak. 101399 do do Jan. 8, 1922 Jan. 24,1922 East St. Louis, Mo. 2 Date approximate. 3 An obvious error; return reported after the close of the hunting season; bird possibly killed illegally. 4 Under direction of Joseph Pulitzer, of St. Louis, Mo. 6 Between Forts Vermilion and Chipewyan. RETURNS FROM BANDED BIRDS, 1920 TO 1923 Mallard — Continued 15 No. Age and Banding record Return record sex Operator Locality Date Date Locality 101962 L. V. Walton _ Cuivreld., Mo.. Jan. 15,1923 Jan. 31,1923 Reelfoot Lake, Tenn. 101997 do _ May 3,1923 katchewan (25 mi. NE.), 102006 cf do do. Jan. 19,1922 Dec. 30,1922 Dec. 28,1922 Meredosia, 111. 102023 cf do do do Apex, Mo. 102028 9 do.. do Jan. 22, 1922 Dec. 6, 1922 Middletown, Iowa. 102036 cf do do _ Jan. 30, 1922 Nov. 17, 1922 Salt River, Ralls County, Mo. 102039 cf Nov. 24, 19222 Jan. 13,1923 Nov. 8, 1922 Appleton, Minn. 102040 do 102047 cf do do.. do Mouth of Missouri River (1 mi. S.). 102049 cf do do do Nov. 24, 1922 Dec. 9, 1922 Rock County, Wis. Grafton, 111. (3 mi. N.). 102050 cf do do ..—.do 102057 9 do do do Apr. 19,1923 Wordsworth, Saskatche- wan. 102067 9 do... do... do Dec. 13,1922 Arkansas River (E. part Okla.). 102070 cf do Feb. 1, 1922 Dec. 20,1922 Nov. 19, 1922 Dee. 10,1922 St. Charles, Ark. 102072 9 do Feb. 12,1922 Feb. 16,1922 Coahoma County, Miss. Webb City, Mo. (16 ml 102075 cf do.. do NW.). 102087 o do do Feb. 21,1922 Latter part of season, Mount Vernon, Ind. 1922. 200103 cf do do Feb. 22,1922 Dec. 23,1922 Sheldons Grove, 111. 200105 cf do do do Dec. 29,1922 Elk City, Okla. (15 mi. S.). 200110 9 . ...do... do ... do Nov. 28, 1922 Nov. 18, 19222 Wagner, Okla. Alta Loma, Tex. 200112 cf do do Mar. 1, 1922 200119 9 Mar. 4,1922 Oct. 16,1922 N. pt. Clearwater Coun- ty, Minn. 200122 cf do do do Nov. 5,1922 Herman, Minn. (3J mi. SE.). 200123 9 Dec. 14,1922 Feb. 3, 1923 Crowley, La. (18 mi. N.) Portage des Sioux, Mo. 8 200124 cf do do do 200136 cf do do.. Jan. 6, 1923 Feb. 13,1923 Do.8 200142 9 Jan. 7, 1923 do Do.8 200142 9 do do Mar. 24, 19232 Feb. 24, 19232 May 19,1923 Malden, Mo. 200179 9 Jacob, El. 200208 Jan. 8, 1923 wan (8 mi. W.). 200263 Jan. 10, 1923 Feb. 25,1923 Aug. 3, 1923 East Cairo, Ky. Lake la Plonge, Saskatch- ewan. 200339 200355 Jan. 11,1923 July 4,19232 Cross Lake Post, Mani- toba. 200373 Jan. 18, 1923 Mar. 16, 1923 May 1,19232 Finley, Tenn. Momence, 111. (5 mi. E.). Sled Lake, Green Lake, 200395 200417 Jan. 12,1923 Saskatchewan (20 mi. NE.). 200586 Jan. 17,1923 Feb. 10,1923 Mar. 3,1923 Feb. 13, 1923 Delaplaine, Ark.. Claryville, Mo. Portage des Sioux, Mo. 9 Lake Winnipegosis, Manitoba. 200590 205251 9 Feb. 2, 1923 Feb. 8, 1923 205321 9 June 5, 1923a 8 Date approximate . 'Trapped by John Broeker and again released. 16 BULLETIN 1268, U. S. DEPAETMENT OF AGRICULTURE Mallard — Continued No. Age and Banding record Return record sex Operator Locality Date Date Locality 205330 d" L. V. Walton. Cuivreld., Mo.. Feb. 9, 1923 Mar. 17, 1923 Lilbourn, Mo. 205342 c? do do. Feb. 13, 1923 205522 d1 do___ Mar. 1, 1923 Mar. 25, 1923 205553 9 do Mar. 2,1923 “Spring, 1923” wan. 205590 c? do.— Mar. 24, 1923 205592 9 do___ May 21, 1923 205787 c? _ ___do Mar. 17, 1923 Apr. 9, 1923 Peruque, Mo. 205808 d1 do Mar. 19, 1923 May 4,1923 Summerberry River, Saskatchewan. 205840 9 do Mar. 20, 1923 Apr. 19,1923 Pinehurst Lake, Al- berta. *36837 9 A. A. Allen Ithaca, N. Y Mar. 18, 1918 Sept. 6,1920 Quill Lake, Saskatche- wan. *36839 o’ do Apr. 18,1918 Fall, 1919 Near Chamberlain, S. Dak. *36840 d Mar. 18,1918 Nov. 11, 1920 Grand Chenier, La. *36845 d Apr. 8, 1918 Nov. 19, 1921 *36858 9 . ___do Mar. 18, 1918 Sept. 25, 1920 Ferguson Flats, Alberta. t*37425 d H. S. Osier.... Lake Scugog, Sept. 25, 1922 Oct. 16,1922 Lake Scugog, Ontario. Ontario. 4616 do _ Sept. 27, 1920 Quitman County, Miss. 4640 c r do do Sept. 30, 1920 Nov. 27, 1920 Wrightville Beach, N. C. 4697 Oct. 23, 1920 Ceasarea, Ontario. 4698 Nov. 18, 1920 Do. 5101 Nov. 6, 1920 Dec. 6, 1920 Long Point Bay, Lake Erie, Ontario. 5103 Jan. 15, 1921 St. Andrews, Fla. 5104 Nov. 15, 1920 Port Rowan, Ontario. 5158 Sept. 12, 1921 Nov. 11, 1922 Toledo, Ohio (2J mi. E.). 5159 do do do Oct. 4, 1921 Durham County, Onta- rio. 101180 do do Aug 22, 1922 Sept. 2,1922 Orillia, Ontario. 1101258 Aug. 29, 1922 Oct. 10,1922 Lake Scugog, Ontario. 101259 Oct. 19,1922 Ottawa County, Ohio. 228442 9 Oct. 7, 1922 Nov. 9,1922 Fort Valley, Ga.(6mi.N.), 228624 Oct. 21,1922 Feb. 10,1923 Blaney, S. C. 202420 im H. L. Felt A.. Findlater, Sas- Sept. 3,1922 Nov. 25, 1922 Cataro, La. (3 mi. N.). katchewan. Mallard X Black Duck: Anas platyrhynchaX A . rubripes 102588 do do. do May 12, 1923 Isle a la Crosse Lake, Saskatchewan (20 mi. NE.). 101427 c? do. do Feb. 21,1922 Mar. 1,19222 Gueydan, La. 101540 J it 3 Varieties 3 Variables 4 Methods of computation 5 Basis of interpretation 9 Correlations among the ear characters 10 Correlations between the ear characters and yield 12 Coefficients of the zero and first orders 12 Coefficients of multiple correlation 15 Discussion l 16 Summary 17 Literature cited 19 Department Bulletin No. 1322. — Some Economic Aspects of Farm Ownership. — Trends and Variations in Some Financial Burdens and Benefits of Farm Ownership in the Spring-Wheat Belt During 25 Years. Illustrated from the History of Selected Farms in Cass County, N. Dak., 1896-1920: Purpose and scope of inquiry 2 Long-time average conditions of ownership 3 Trends in ownership conditions , 7 Deviations from trends in ownership conditions 15 Anticipations of the future by owners 16 Adjustments in renting and purchasing farms 22 Summary 23 Department Bulletin No. 1323. — Citrus Pectin: Purpose of investigation 1 Results of previous investigations 2 Methods for determination of pectin 2 Preliminary experiments 3 Methods of extraction 5 Preparation of concentrated solutions and powdered pectin 8 Composition of apple, lemon, and orange pectins 12 Proportions of acid, sugar, and citrus pectin necessary to produce jellies 12 How to make citrus jellies and marmalades 15 Summary 16 Bibliography 16 10 DEPARTMENT OF AGRICULTURE BULS. 1301-1325 Page Department Bulletin No. 1324. — The Oviposition Response op Insects: Introduction 1 Internal physiological condition affecting oviposition 2 Nutrition 2 Age 2 Fertility 2 Internal periodicities 3 External influences affecting oviposition 3 Temperature 3 Humidity 4 Light 5 Air and water currents 6 Surfaces 6 Odorous substances 8 Contact with chemical substances 10 Discussion 10 Conclusions 13 Literature cited 13 Department Bulletin No. 1325. — -Marketing Onions: Development of onion production 1 Classes and types 2 Commercial onion regions 4 Harvesting, grading, and packing 11 Financing the crop 16 Local sales methods 18 Cooperative marketing 20 Storage 21 Transportation 24 Seasonal movement 26 Areas of distribution 26 Wholesale marketing 31 Features of city markets 33 Market preferences 42 Supplies of large cities 42 Seasonal consumption of cities 44 Volume of shipments 44 Price tendencies 46 Forecasting the market 49 Market information 50 Cost of marketing 51 Imports 52 Exports 55 Summary 56 Tables of statistics 57 Onion bulletins and circulars 69 i UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1303 Washington, D. C. V April, 1925 THE PECAN NUT CASE-BEARER* 1 By John B. Gill, Associate Entomologist, Fruit Insect Investigations, Bureau of Entomology CONTENTS Page Introduction 1 History 1 Distribution 2 Food plants 2 Character of injury 3 Description 3 Egg 3 Larva 4 Pupa 4 Adult 4 Page Seasonal history and habits 4 Adult and egg stages 5 Larva stage.. 8 Pupa stage 8 N umber of generations 11 Parasitic insects 11 Control measures 11 Literature cited 12 INTRODUCTION For a number of years the Bureau of Entomology maintained a field laboratory in the southeastern portion of the pecan belt at Monticello, Fla., for the investigation of the biology and habits of pecan insects and to determine the most effective control measures for the more important species under orchard conditions. The results of the greater portion of these investigations have been published (J, 5), 3 but as a part of this work this bulletin brings together the results of a rather detailed study of the life history of the pecan nut case-bearer ( Acrobasis hebescella Hulst), an insect which is now recog- nized as a most formidable enemy of the pecan because of its nut- feeding habits. HISTORY The pecan nut case-bearer was first described by Hulst (7) in 1890 under the name of Acrobasis hebescella. In 1893 Ragonot (9) treated this species, and in 1901 it was listed and figured by Hampson (6), but the figure given is a poor reproduction of the moth. Dyar’s “List of North American Lepidoptera” (/) contains this species, the distribu- tion being given as New Jersey, Texas, Illinois, and Wisconsin. In 1902 E. P. Stiles (11) gave an account of injury in west-central Texas caused by what he called the pecan huskworm, but doubtless duo to Acrobasis hebescella and a closely related species, A. caryivorella Rag. 1 Acrobasis hebescella Hulst; order Lepidoptera, family Pyralidae. 1 Keference is made by number (italic; to “ Literature Cited,” p. 12. 0382° — 26f 1 2 BULLETIN 1303, U. S. DEPARTMENT OF AGRICULTURE Sanderson (10) reported serious injury to the pecan crop by A. caryae in 1903, but it appears to the writer that the greater portion of the damage, if not all, should have been accredited to the species of Acrobasis discussed in this bulletin and to A. caryivorella. Dyar in 1908 (2), in a paper on the species of Acrobasis, includes this insect. In 1914 the writer ( 3 ) made a brief mention of this species in an article on the pecan leaf case-bearer ( Acrobasis nebulella Riley), and in Farmers’ Bulletin 843 (J), appearing in 1917, it is treated at length. The following year Turner (12) and Matz (8) gave general accounts of the insect. DISTRIBUTION Hulst (7), in his original description of Acrobasis hebescella, gave New Jersey and Texas as its habitat; Dyar ( 1 ) listed it from New Jersey, Texas, Illinois, and Wisconsin, and in an article appearing in 1908 (2) mentioned Brownwood, Tex., and East River, Conn., as localities in which it has been collected; and Turner (12) reported it from Thomasville, Ga., and Cairo, Ga. In the United States National Museum collection there are specimens from Brownwood, Tex.; Goodman, Miss.; Bon Ami, La.; and Monticello, Fla. Besides the localities mentioned, the records of the writer show that this species also occurs at Waukeenah, Fla.; Baconton, Dewitt, Putney, and Moultrie, Ga.; Mobile and Fowl River, Ala.; Ocean Springs and Pecan, Miss.; Keithville, La.; and Marshall, Victoria, and Bend, Tex. The pecan nut case-bearer, along with another nut-feeding species of Acrobasis, A. caryivorella Rag., is rather generally distributed in the pecan-growing sections of the Southwest, especially in Texas, where it is reported annually as causing much damage. In the Southeast, however, it has been recorded only from scattered locali- ties in Florida, Georgia, Alabama, Mississippi, and Louisiana. For the last 9 or 10 years this insect has been causing considerable loss to pecan growers in the vicinity of Monticello, Fla., and Thomasville, Ga., and during 1922 it was reported for the first time as occurring in injurious numbers in the large commercial orchards in the Albany, Ga., district, which is at present rated as the most important section for the production of cultivated pecan nuts. In 1923 and 1924 the insect caused very heavy damage in the large orchards at Baconton, Dewitt, Putney, and Albany, all in this district. This species appears to be gradually extending its range of destructiveness, and sooner or later it will probably prove a most formidable pest through- out the greater portion of the pecan belt. FOOD PLANTS The only food plant on which this insect has been found by the writer is the pecan (Hicoria 'pecan), but it is to be supposed that it subsists also on other species of the genus Hicoria. In literature oak and pecan are given as food plants. Dyar (2) states that six speci- mens from Brownwood, Tex., were bred at the insectary of the U. S. Department of Agriculture from larvae on pecan mining into the young buds. In his original description Hulst (7) has the follow- ing statement : A specimen from New Jersey received by Prof. J. B. Smith, has on it a label marked “on oak, Jersey pines, June.” The pin is thrust through an oval close cocoon which was undoubtedly made at or under the surface of the ground. THE PECAH HUT CASE-BEAKER 3 He further states : I have myself found the larval cases of a Phycitid in southern New Jersey on a large-leaved oak, which may be the cases of this species. They were horn- like, much resembling those of indigenella. The larva turned over the edge of the large leaf binding the edges, and forming a habitation large enough to move about freely within. The case itself was fastened within with threads of silk. The habitat given by Hulst is New Jersey and Texas, but there is no mention of the food plant for the material from Texas, which was probably taken on the pecan. In the writer’s opinion, the record of the New Jersey specimen “from oak” is very doubtful, because the description of the cocoon and the manner of pupation do not coincide with the facts regarding Acrobasis hebescella as determined in an extensive study of this species. CHARACTER OF INJURY The larvae of the pecan nut case-bearer may attack either the tender shoots or the immature nuts. The larvae that pass the winter in hibernacula around the buds cause damage in early spring by attack- ing the tender shoots, in which they tunnel and eat out the interior, leaving the outside intact. Many of the attacked shoots wilt and turn brown, and others are broken off by the winds (PL IV) . Such injury is not very serious, compared with the damage caused by the first and second brood larvae, which confine their attacks to young green nuts. During May the first-brood larvae make their appearance and bore into the recently set nuts. At the point of attack pellets of frass or borings are cast out (PI. II) and held together by means of fine silken threads that form a short silk-lined tube. Nuts injured by this insect always show the characteristic mass of frass protruding from the place where the larvae gained entrance, which is invariably in the side of the nut near the basal end. The larvae of the second generation attack the nuts in the same manner as those of the first generation, but the damage to the crop is not so extensive because of the size of the nuts at the time of attack (PI. Ill) . Early in the season a single larva may destroy several nuts before attaining full growth, while later in the season one or two nuts seem to be suf- ficient for its subsistence. The writer’s observations show that by far the greatest damage to the nut crop is inflicted by the larvae of the first generation. The larvae of the third generation, which make their appearance in the late summer, usually feed very little. They seem to prefer the shucks or hulls, in which they gnaw only through the surface, forming small, narrow, elongated tunnels of frass particles of a rather delicate or flimsy texture. Such injury does not interfere with the normal development of the nuts. Some of these larvae also feed slightly on the leaf petioles and succulent shoots, but the damage thus caused is insignificant. When the larvae of this generation seek hibernation quarters they are usually a little less than one-tenth of an inch long. DESCRIPTION EGG The egg (PI. I, A) is elliptical, convex above and flattened below, with the surface finely wrinkled. When first deposited the egg is greenish white, and as incubation advances it takes on a reddish 4 BULLETIN 1303, U. S. DEPARTMENT OF AGRICULTURE tinge. The eggs average 0.56 millimeter in length by 0.34 milli- meter in width. The moths deposit their eggs singly on the calyx end of the nut or on the side of the nut under the calyx lobes. LARVA The larva (PL I, B at right, and D) is nearly cylindrical, tapering slightly toward each end, but more posteriorly than anteriorly. Eight full-grown larva when extended averaged 14.59 millimeters in length by 2.08 millimeters in greatest width. The head and mouth parts are dark brown, and the prothoracic shield is pale brown, bisected by a rather inconspicuous faint whitish area. The general color of the body is dirty olive green, darker dorsally and laterally than ventrally. The body is sparsely covered with fine whitish hairs, the skin being quite wrinkled, especially in the thoracic region. The thoracic legs are about the same color as the ventral part of the body, and each terminates in a single brown claw. There are five pairs of prolegs and the anal pair is about twice as long as the others. PUPA The pupa (PI. I, B, at left) is of the usual form and wdthout striking markings. When first formed it has a decided olive-green cast, but with age it changes to a fight brown. The abdominal segments are finely punctate, the anterior portion of the last segment having a dark brown transverse band dorsally. The anal extremity is armed wdth a short, sharp brownish tooth, directed nearly at right angles to the body axis, and four slender light-brown hooked spines, which arise from the lower part of the tip of the last segment and project forward. The size of the pupa is decidedly variable, the average measurements for six individuals being 8.1 millimeters by 2.2 millimeters. The pupa is formed inside the attacked shoot for the spring brood (PI. IV), and within the infested nut for the first and second broods (PI. I, E) . Upon issuance of the moth the pupal skin is not extended. ADULT (PL I, C) The pecan nut case-bearer was first described in 1890 by Hulst (7) under the name Acrobasis Jiebescella. The original description is as follows : Expands 18-20 mm. Labial palpi blackish gray. Head ochreous fuscous, thorax dark fuscous. Abdomen ochreous gray, annulate with fuscous. Fore wings short, very broad, strongly arched on costa and inner margin, ochreous fuscous, quite dark; lines indistinct, basal hardly discernible, faintly gray, edged outwardly with black at costa; scale ridge black, short; outer line dentate, shown by black border lines on ground color; discal spots quite distinct, confluent. Hind wings dark even fuscous. The measurements of moths bred from pecans varied from 14.5 to 19 millimeters across the expanded wings. SEASONAL HISTORY AND HABITS All records given for the various stages of the pecan nut case-bearer were obtained at Monticello, Fla. In the rearing work pertaining to this insect glass jars were used as cages, which were kept in an open- air insectary. In the discussion of the fife history of this species, a “generation” is considered to begin with the egg stage and to end with the adult or Bui. 1303, U. S. Dept, of Agriculture Plate I The pecan Nut Case-Bearer A, Eggs on nut. Greatly enlarged. B, Pupa at left larva at risht- E parent 1 E Infested moth. Enlarged. D, Immature pecan nut infested by larva. Slightly enlarged. L, iniesicu pecan nut showing place of pupation. Slightly enlarged Bui. 1303, U. S. Dept, of Agriculture Plate II Immature Pecan Nuts, Showing Injury. Natural Size Bui. 1303, U. S. Dept, of Agriculture Plate III Clusters of Infested Pecan Nuts. Natural Size Bui. 1303, U. S. Dept, of Agriculture Plate IV Pecan Shoots, Showing Infestation by Larvae in the Spring. Reduced THE PECAN NUT CASE-BEARER 5 moth stage. The term “ brood7’ is used in speaking collectively of the individuals of any stage of a particular generation, as egg, larva, pupa, or adult. The overwintering larvse, after transforming in the spring to pupae and in turn to adults are referred to as “ spring-brood pupae ” and “ spring-brood moths.” Although the spring-brood pupae and moths are the first to make their appearance during the growing season, they are not designated as first-brood pupae and first-brood moths, but these terms are reserved for the pupae and moths of the next succeeding generation. ADULT AND EGG STAGES TIME OF EMERGENCE OF THE SPRING-BROOD MOTHS From material under observation during the season of 1916 it was determined that spring-brood moths emerged from May 7 to May 23, and in 1917 the period of emergence was May 2 to May 17. Moths in the pecan orchard evidently came forth about the same time as those in rearing cages, a fact which was determined by numerous field observations. There is, however, one record of collecting an empty pupal skin in the field as early as May 1, 1917, indicating that this particular individual emerged earlier than the first insectary rearing. The dates of issuance of 60 individuals are shown in Table 1. Table 1. — Time of emergence of spring-brood moths of the pecan nut case-bearer , Monticello, Fla., 1916 and 1917 1916 1917 Num- Num- Num- Num- Date of emergence ber of Date of emergence ber of Date of emergence ber of Date of emergence ber of moths moths moths moths 1 5 May 2 1 2 8 1 17 5 3 5 14 1 10 1 18 3 4 3 16 3 11 2 19 3 5 1 17 3 12 1 20 1 1 13. 5 23 . . 1 10 4 24 14 1 15 6 Total 36 TIMES OF EMERGENCE OF FIRST AND SECOND BROOD MOTHS Records on the times of emergence of first and second brood moths are given in Table 2 for 1914, 1915, 1916, and 1917. From material under observation during the season of 1914, it was determined that first-brood moths emerged from June 11 to July 2, covering a period of 22 days. The maximum emergence occurred on June 23, when 15 moths came forth. The second-brood moths began to appear on July 23 and continued emerging until August 26, a period of 35 days for all individuals under observation. Probably the times of emergence of both the first and second brood moths varied some- what from the dates based on reared specimens, but no doubt the data given closely approximate the actual times of emergence under natural conditions. The records for 1915 show that the first-brood moths emerged from June 18 to July 15, the maximum emergence of 13 individuals occur- 6 BULLETIN 1303, IT. S. DEPARTMENT OF AGRICULTURE ring on June 24 and emergence for all moths extending over a period of 28 days. The earliest record of emergence for the second-brood moths occurred August 2 and emergence continued until August 31, a period of 30 days. , The emergence of the first-brood moths for 1916 began June 17 and lasted until July 16, covering a 30-day period. For the second-brood moths emergence took place from August 3 to August 30, a period of 28 days. For 1917 the first-brood moths were found to emerge from June 13 to July 12, covering a period of 30 days. The greatest number to appear on a single day was 16 on June 29, while June 23 to July 2 marked the period of maximum emergence. The second-brood moths emerged from August 2 to August 30, a 29-day period for emergence. Table 2. — Times of emergence of the first and second brood moths of the pecan nut case-bearer, Monticello, Fla., 1914, 1915, 1916, and 1917 First brood Second brood Date 1914 1915 1916 1917 Date 1914 1915 1916 1917 Number Number Number Number Number Number Number Number 2 July 23 1 12 2 24 __ 3 13 1 1 25 4 14 6 26 4 15 5 1 27 4 16 5 1 28 9 17 2 1 1 29 5 18 2 2 4 2 30 9 19 14 2 3 31 5 20 6 2 4 1 Aug. 1 3 21 8 2 4 2_ 4 1 1 22 7 4 3 3 3 9 1 1 23 15 2 5 10 4 10 1 1 24 12 13 5 12 5 6 2 25 10 10 2 6 6 1 2 26 5 5 8 14 7 6 6 1 27 . 8 5 5 6 8 8 1 2 28 2 4 6 9 9 1 11 29 2 3 3 16 10 8 7 1 30 1 2 7 5 11 10 5 1 1 1 1 6 7 12 12 10 3 ' 2 2 1 10 14 13 6 6 2 3 3 2 8 1 14 4 4 a 4 4 9 4 15 9 3 1 1 5 9 2 1 16_ 1 1 1 6 3 3 17 5 7 3 1 7 l 1 4 18 4 3 6 1 8 2 19 4 9 5 1 9 2 1 2 20 8 8 3 10 21 5 2 1 11 1 3 22 7 1 1 12 1 23 4 2 a 13 __ 1 1 24 2 4 1 i 14 25 3 i 15 2 26 1 3 2 2 16 1 27 1 2 2 28 1 2 2 29 1 30 1 1 31 1 Total. 112 70 100 141 Total 166 116 42 30 LENGTH OP LIFE OF MOTHS The length of life of moths varied from 2 to 15 days, the average for those under observation being about 6 days. The data bearing on this phase of the life history are insufficient to make any general- ization, as only a few observations were made to determine this point. THE PECAN NUT CASE-BEAKER. 7 OVIPOSITION AND LENGTH OF INCUBATION PERIOD It was very difficult to get moths to oviposit in confinement. From material under observation it was determined that a period of from 2 to 7 days elapsed from the time of emergence to the time of first oviposition. The moths confined in cages usually deposited all eggs within a period of 3 days. In confinement moths deposited eggs indiscriminately upon the nuts, leaf petioles, foliage, and occa- sionally on the sides of the glass breeding jars. It seems that under natural conditions egg deposition invariably takes place on the nuts at the calyx or on the side immediately under the calyx lobes. As is shown in Table 3, the average length of the period of incu- bation of 56 first-brood eggs was found to be 9.48 days, the maximum being 10 and the minimum 7. Table 3. — Length of incubation period, of first-brood eggs of the pecan nut case- bearer, Monticello, Fla., 1917 Number of eggs from which larvae emerged Date of oviposition Date of hatching Length of incubation period 29 May 11... May 21... Days 10 26 13... 22... 9 1 15... 22... 7 The length of the incubation period for second-brood eggs is given in Table 4. Table 4. — Length of incubation period of second-brood eggs of the pecan nut case- bearer, Monticello, Fla., 1916 and 1917 Number of eggs from ■which larvae emerged 1916 Date of oviposition June 26- do.. J une 27 . June 28. July 3,-. do.. July 14-. do.. do.. Date of hatching July 1-- July 2.. July 3_. July 4.. July 8_. July 10. July 18. July 19. July 20. Average for 47 individuals. Length of incu- bation period Days Number of eggs from which larv» emerged Date of oviposition July I.. do. July 5.. do. do. July 6.. do. July 7.. Average for 136 individuals Date of hatching Length of incu- bation period July 7_. July 8.. July 10. July 11. July 12. do. July 13. do. . Days 6. 17 Observations on a limited number of third-brood eggs hatching in August showed that the period of incubation ranged from 4 to 6 days, the average being 4.95 days. 8 BULLETIN 1303, U. S. DEPARTMENT OE AGRICULTURE LARVA STAGE EMERGENCE OF LARVAE FROM HIBERNATION The overwintering larvae become active and leave their winter cases (hibernacula) during the latter part of March or the first part of April, at which time the buds on pecan trees begin to open. After feeding slightly upon the unfolding buds, these spring-brood larvae migrate to the rapidly growing succulent shoots, in which they feed by tunneling out the interior, keeping the burrows open by casting out the frass pellets from the hole where the initial attack was made (PI. IV). It has been determined from field observations that the larvae usually begin to attack the pecan shoots during the second week in April, and after feeding in this manner for two weeks or more the larvae transform to pupae within their burrows. During 1916 the spring-brood larvae that were under observation pupated from April 24 to May 12, and in 1917 from April 19 to May 1. Records show, however, that a few larvae can be found on pecan trees, espe- cially the Stuart variety, as late as the second week in May. FIRST-BROOD LARVA! The length of larval life for the first brood ranges from 22 to 29 days, the average being about 26 days. The maximum hatching of first-brood larvae occurs during the third week in May, but the period of hatching is quite extended. SECOND-BROOD LARVA! The majority of the second-brood eggs hatch during the last week in June and the first 10 days of July, but the hatching for all eggs of this brood extends over a long period. The average length of the larva stage for the second brood was found to be 25.27 days. THIRD-BROOD LARVA! The hatching of third-brood larvae begins about the middle of August and continues until the first part of September. The latest hatching of third-brood larvae in rearing cages occurred on September 3, but under natural conditions it is likely that larvae hatch over a longer period. The third-brood larvae feed very sparingly for three or four weeks and then go into hibernation by constructing hiber- nacula, which are attached to the buds. Although the larvae feed for several weeks, they do not attain a size greater than one-tenth of an inch. In rearing work it has been observed that the larvae are very prone to construct their hibernacula prematurely when the food plant in breeding cages is not kept in the best of condition. The larvae remain in their hibernacula throughout the winter, and with the advent of spring they become active just as the buds on the pecan trees are unfolding. PUPA STAGE PLACE OF PUPATION OF THE OVERWINTERING LARVA! The spring-brood pupae are formed within the shoots in which the larvae complete their growth. Near the mouth of its burrow the full- grown larva prepares out of particles of frass and excrement a flimsy silken-lined cell or cocoon, where its transformation to the pupa stage occurs (PI. IV). THE PECAN NUT CASE-BEARER 9 TIME OF PUPATION OF OVERWINTERING LARVAE AND LENGTH OF SPRING-BROOD PUPAL PERIOD Investigations show that the time of pupation of the overwintering larvae is variable. Records for 1916 give April 24 as the earliest date of pupation, and the latest as May 12; while for 1917 pupae were forming in rearing cages from April 19 to May 1. During the season of 1917 overwintering larvae were observed in the field after May 1, showing that insectary records do not actually indicate the complete period at which transformation to pupae occurs. Before making any generalization on the time of pupation, more extensive data are de- sirable, as observations have been confined to a limited number of individuals. During 1916 and 1917 the length of the pupal period of the spring brood was determined for 19 individuals, as shown in Table 5: Table 5. — Length of pupal period of spring-brood pupse of the pecan nut case- bearer, Monticello, Fla., 1916 and 1917 1916 1917 Num- ber of Date of indi- pupation viduals Date of emergence Days as pupa Num- ber of indi- viduals Date of pupation Date of Days as emergence pupa 1 1 1 1 1 2 2 1 1 1 Apr. 24. Apr. 28- May 2.. May4. . May 5. . May 6.. May 7.. do.. May 8— May 12. May 11. May 13. May 14. May 15. May 16. May 17. May 18. May 19. do.. May 23. Average for 12 individuals. 17 15 12 11 11 11 11 12 11 11 12 1 1 1 1 1 1 1 Apr. 19. Apr. 20. Apr. 28. do.. Apr. 29. Apr. 30. May 1. May 2_. May 3. May 14. May 16. May 17. do_. May 16. 13 13 16 18 18 17 15 Average for 7 individuals. 15.85 TIME OF PUPATION AND LENGTH OF PUPAL PERIOD OF THE FIRST AND SECOND BROOD PUP.E Of the first-brood larvae under observation that transformed to pupae during 1914 the first pupated June 9, and during 1916 and 1917 the first pupae appeared June 5 and June 2, respectively. Pupa- tion continued in 1916 until June 29 and in 1917 until July 2, while in 1914 the last pupa for insectary material was recorded on June 19. It should be stated, however, that larvae under natural conditions were probably pupating for a week or 10 days later than pupation records for 1914 indicate, as it was determined on June 20 from field- collected material that only 85 per cent of the larvae had pupated. These records are given in Table 6. The second-brood larvae transform to pupae during the latter half of July and the greater part of August. The time oi pupation for a number of individuals of the second brood is also given in Table 6. These records, however, do not represent the entire period of pupation for the second brood, as only a limited number of trans- forming larvae were under observation during some seasons. 10 BULLETIN 1303, U. S. DEPARTMENT OF AGRICULTURE Table 6. — Length of 'pupal period of the first and second brood pupse of the pecan nut case-bearer, Monticello, Fla., 1914, 1915, 1916, and 1917 Num- ber of Date of Date of Days as indi- pupation emergence pupa viduals 1914— Fiest Brood June 9__ June 10. June 12_ do_- June 13. June 15_ do._ June 16_ June 17_ do__ June 18- June 19. June 19. June 22_ June 21_ June 23. June25- June24. June 26- June 27. do_- June 28- do_. June 29. Average for 22 individuals. 1914— Second Brood l July 23 10 1 11 2 do July 25 12 1 10 2 11 3 July 26 12 2 11 2 do ___ July 27 12 1 July 16 _ _ do 11 1 12 4 July 18 10 1 July 19. Julv 29 10 1 July 20. 9 1 do July 30 10 3 July 31 11 2 July 21 ___ _do 10 1 July 23. Aug. 2 . . 10 1 do Aug. 3 11 1 July 27 11 1 July 28 11 1 July 29.. Aug. 9 --_ 11 1 do Aug. 11 13 1 Aug. 3 . . _ . Aug. 13. 10 1 Aug. 5. 11 1 12 1 Aug. 12 .. 12 1 Aue. 13 13 A verage for 39 individuals 11 1915— Second Brood 1 July 25 Aug. 3. -_ . 9 1 do Aug. 4__ 10 1 July 26 . Aug. 5 10 2 July 31 Aug. 10_ _ .. 10 3 do 11 4 12 1 do 13 4 Aug. 1 11 2 do Aug. 13 12 6 Aug. 2__ . do 11 1 do Aug. 14 12 1 12 2 9 3 Ido Aug. 14. 10 Average for 32 individuals 10.91 1916 — First Brood June 5 June 7 _ do June 19 1 June 8 2 j une 9 1 1 June 11 1 June 17 1 1 2 July I... 1 July 2 1 2 June 21 1 1 do July 5 1 June 26 July 7 1 __ __do July 8... 1 June 27 do ___ 1 June 28 - July 9 1 June 29 _ _ Julv 11 Average for 24 individuals Num- ber of indi- viduals Date of pupation Date of emergence Days as pupa 1916 — First Brood— Continued 1916 — Second Brood Aug. 8_. do... do... Aug. 9. - Aug. 10 _ Aug. 15- do... Aug. 17- Aug. 17. Aug. 18. Aug. 19. do_. Aug. 20 _ Aug. 24 _ Aug. 26. Aug. 28- Average for 11 individuals. 1917— First Brood June 2 June 3 June 7 June 8 June 9 June 12_ do June 13 June 14 do June 15 do June 16 June 17 do June 18 June 20 June 21. June 23 June 25 do June 29 July 2 June 13. June 14.. June 18. June 21. do... June 24. June 25. June 26. do.. June 27. do.. June 28. do... do— . June 29. do... July 2... do... July 4... July 6... July 7... July 11- July 12.. Average for 32 individuals. 1917— Second Brood 2 Aug. 1 Aug. 13- 2 Aug. 2 Aug. 14. 1 Aug. 14 Aug. 27. 1 Aug. 19 Aug. 30- Average for 6 individuals. 11.84 12 THE PECAN NUT CASE-BEARER 11 The data on the length of the pupal period of the first-brood pupae for 1915 are incomplete, but in the case of the few indi- viduals under observation the pupal period lasted from 10 to 11 days. NUMBER OF GENERATIONS Life-history studies have shown that under normal conditions the pecan nut case-bearer has three generations annually, at Monticello, Fla., and that it passes the winter in the immature larva stage of the third generation. These third-brood larvae after feeding sparingly for a few weeks migrate during the latter part of September or the first part of October to the buds, where they construct small, com- pactly woven hibernacula, in which they hibernate until the following spring, when the buds on pecan trees are beginning to open. In con- finement it may be possible for the pecan nut case-bearer to have four generations. PARASITIC ENEMIES The writer has on several occasions reared a number of parasitic insects from the larvae and pupae of the pecan nut case-bearer as follows: Exorista ( NemoriTla ) pyste Walk.,3 Habrobracon variabilis Cush.,4 CalliepJiialtes grapholithae Cress.,4 Cremastus ( Zaleptopygus ) sp.,5 and Angitia sp.4 The most effective parasite is the tachinicl fly Exorista pyste Walk., which was reared in large numbers from the larvae and pupae, and is no doubt a very important factor in the natural control of this pest. The braconid Habrobracon variabilis also was frequently bred, and is perhaps second in importance among the parasites of the pecan nut case-bearer. CONTROL MEASURES Investigations so far conducted show that the best method of control against the pecan nut case-bearer is spraying with arsenate of lead.6 The arsenate should be used at the rate of 1 pound of the powdered form, or 2 pounds of the paste form, to each 50 gallons of water, to which should be added the milk of lime from 3 pounds of slaked lime. The addition of lime is necessary to prevent arsenical injury to the foliage and nuts. Three applications will be required and should be at the following periods : 1. Soon after the nuts have set, at which time they are about the size of garden peas. 2. One week or 10 days later. 3. From four to five weeks after the second application. The date for the first application at Monticello, Fla., during the season of 1915 was found to be May 15, and in 1916 it was May 12. The time of spraying, however, will vary somewhat according to latitude and the character of the season. The first and second appli- cations are the most important in the control of the pecan nut case- bearer, as most of the damage to the nuts is usually caused by the larvae of the first generation. wSince only a small portion of the nut crop is attacked by the second-brood larvae, it is suggested that if growers desire to reduce spraying of their orchards to a minimum 3 Determined by W. R. Walton. 4 Determined by It. A. Cushman. 1 Determined by It. A. Cushman. The genus Cremastus now embraces the old genus Zaleptopygus. 9 In connection with life-history investigations, spraying experiments were conducted during 1915 and Kite, the results of which were reported in detail in Farmers’ Bulletin 843 of the United States Department of Agriculture, pages f> to 9. Further spraying experiments were begun in the spring of 1917, but could not he successfully carried out on account of the mildness of the infestation of the insect in the orchard selected for this work . . 12 BULLETIN 1303, U. S. DEPARTMENT OF AGRICULTURE the last application be omitted, but that the first and second appli- cations be made very thorough. It is realized that the control recommendations given for this insect are not altogether satisfactory, but in the light of present knowledge they are the best that can be offered. Further detailed investiga- tions are now being made to determine, if possible, a more effective method of controlling the pecan nut case-bearer, as well as other pecan insects of economic importance. Some time must necessarily elapse before reporting on these investigations, and in the meanwhile pecan growers are urged to use the control measures recommended. LITERATURE CITED (1) Dyar, H. G. 1902. A list of North American Lepidoptera and key to the literature of this order of insects, fiul. U. S. Nat. Mus., no. 52, 723 p. Page 419: Listed from New Jersey, Texas, Illinois, and Wisconsin. (2) 1908. Notes on the species of Acrobasis, with descriptions of new ones. In Proc. Ent. Soc. Wash., v. 10, no. 1-2, p. 41-48. Page 44: Makes mention of rearing six specimens from pecans from Brownwood, Tex.; one specimen from East River, Conn., host not being given. (3) Gill, J. B. 1914. The pecan case-bearer. In Proc. Fla. Hort. Soc. for 1914, p. 148- 150. (4) Page 150: Brief mention of injury by Acrobasis hebescella. 1917. Important pecan insects and their control. U. S. Dept, of Agr., Farmers’ Bui. 843, 48 p., 58 fig. Pages 3-9: General account of this species, giving description of stages, life history notes, and control measures. 1917. The pecan leaf case-bearer. U. S. Dept. Agr. Bui. 571, 28 p., 3 pi. (6) Hampson, G. F. 1901. Supplement au tome premier de la Monographie des Phycitinae [for E. L. Ragonot], In Romanoff, N. M., Memoires sur les Lepidopteres, v. 8, p. 511-559. Paris. Page 520, pi. 50, fig. 10: Lists and figures moth; not a good reproduction of the subject. (7) Hulst, G. D. 1890. The Phycitidae of North America. In Trans. Amer. Ent. Soc., v. 17, p. 93-229, pi. 6-8. Page 126: Original description of Acrobasis hebescella. Habitat given, New Jersey and Texas. (8) Matz, J. 1918. Diseases and insect pests of the pecan. Fla. Agr. Exp. Sta. Bui. 147, p. 135-163, fig. 45-73. Page 152: Brief account. (9) Ragonot, E. L. 1893. Monographie des Phycitinae et des Galleriinae. In Romanoff,, N. M., Memoires sur les Lepidopteres, v. 7-8, 658+602 p., 57 pi. St. Petersburg and Paris. Volume 7, p. 109: Brief account. (10) Sanderson, E. D. 1904. Insects of 1903 in Texas. In Proceedings of the sixteenth annual meeting of the Association of Economic Entomologists, U. S. Dept. Agr., Div. Ent., Bui. 46, p. 92-96. Page 95: Reports serious injury to pecan crop of Texas by Acrobasis caryae, but it appears likely that most of the damage should have been accredited to -4. hebescella. (11) Stiles, E. P. 1902. Pecan huskworm. In Farm and Ranch, v. 21, no. 50, p. 10-11. Gives account of injury in west-central Texas caused by what he calls the pecan husk -worm, but doubtless due to the pecan nut case-bearer. (12) Turner, W. F., Spooner, C. S., and Crittenden, C. G. 1918. Pecan insects, pecan scab, and pecan diseases other than scab. Ga. St. Bd. Ent., Bui. 49, 48 p., 15 pi. Pages 14-19: Turner gives general account of the species based on work conducted at Thomasville, Ga. WASHINGTON : GOVERNMENT PRINTING OFFICE : 1925 UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1313 Washington, D. C. January 26, 1925 FUMIGATION AGAINST GRAIN WEEVILS WITH VARIOUS VOLATILE ORGANIC COMPOUNDS By IRA E. NEIFERT, F. C. COOK, R. C. ROARK, and W. H. TONKIN Insecticide and Fungicide Laboratory, Miscellaneous Division, Bureau of Chemistry and E. A. BACK and R. T. COTTON, Bureau of Entomology CONTENTS Page Purpose of Investigation 1 Experimental Procedure 2 Effect of Volatile Organic Compounds on Weevils 3 Relation between Volatility and Toxicity of Fumigants 19 Effect of Fumigation on Weevils in the Pres- ence of Grain 24 Fire Hazard from Fumigation 28 Page Effect of Fumigation on Milled and Baked Products 34 Additional Fumigation Tests with Ethyl Acetate and Carbon Tetrachloride. 35 Effect of Ethyl Acetate-Carbon Tetrachloride Fumigation on Germination of Seeds ... 38 Summary 38 Literature Cited 39 WASHINGTON GOVERNMENT PRINTING OFFICE 1925 UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1313 Washington, D. C. ▼ January 26, 1925 FUMIGATION AGAINST GRAIN WEEVILS WITH VARIOUS VOLATILE ORGANIC COMPOUNDS By Ira E. Neifert, F. C. Cook,* 1 R. C. Roark, and W. H. Tonkin, Insecticide and Fungicide Laboratory, Miscellaneous Division, Bureau of Chemistry, and E. A. Back and R. T. Cotton, Bureau of Entomology 2 CONTENTS Page Purpose o f investigation 1 Experimental procedure 2 Effect of volatile organic compounds on weevils 3 Relation between volatility and toxicity of fumigants 19 Effect offumigation on weevils in the presence ofgrain i 24 Fire hazard from fumigation 28 Effect of fumigation on milled and baked products 34 Additional fumigation tests with ethyl acetate and carbon tetrachloride 35 Effect of ethyl acetate-carbon tetrachloride fumigation on germination of seeds 38 Summary 38 Literature cited 39 PURPOSE OF INVESTIGATION Weevils destroy many million dollars’ worth of wheat and other grains annually. Carbon disulphide is extensively used as a fumigant against these insects, but, although efficacious, it has serious disad- vantages. It has an extremely disagreeable odor and in moderate concentrations its vapor is poisonous to man. Although carbon disulphide is volatile, millers occasionally complain that wheat which has been treated with it retains its odor, and it has been shown that the baking quality of flour from carbon disulphide fumigated wheat is sometimes injured (8).s The really serious objection to the use of carbon disulphide as a fumigant, however, arises from the fact that it is readily inflammable and that its vapor when mixed with air is highly explosive. For this reason fire insurance companies refuse to carry the fire risk on elevators during the time carbon disulphide is being used to treat the grain contained in them. Even more important is the action taken by the General Managers’ Association of Chicago, representing the leading railway systems of the United States, in adopting the following resolution: That because of the highly inflammable character of solution of carbon disul- phide and tetrachloride, its use for fumigating cars to be loaded with grain be prohibited, except that the Illinois Central may continue its use at New Orleans 1 Deceased June 19, 1923. 1 Credit is due Ma]. A . Gibson, formerly of the Chemical Warfare Service, for obtaining poisonous mate- rials from thelaboratoriesofEdgewood Arsenal; G. W. Kirby, of the Insecticide and Fungicide Laboratory, Bureau of Chemistry, for assistance in the chemical work; J. II. Cox, Harold Anderson, and the Milling and Baking Laboratory of the Grain Division, Bureau of Agricultural Economics, for assistance in carrying out the car-fumigation tests. a Italic numbers in parentheses refer to Literature Cited, p. 39, 16686°— 2ot 1 2 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE and the Baltimore & Ohio at Locust Point, Md., when cars are isolated and protected, until some other satisfactory suitable substance can be provided, and further, that each carrier, member of this association, shall issue, without delay, necessary instructions prohibiting its use. The following resolution was also adopted at the same time by this association : That because of the presence of bran bug and weevil in grain and the great danger being done thereby, the chairman of this association communicate with the Bureau of Agriculture, Washington, D. C., advising of the action taken by this association and recommending that such investigations as may be necessary be made by the Government to produce a substance for this purpose which can be used with safety. The investigation reported in this bulletin was undertaken because of the action taken by these railroad officials. Carbon tetrachloride, hydrocyanic acid gas, sulphur dioxide, chloropicrin, naphthalene, phosgene, arsine, cyanogen chloride, and many other substances have been tested as fumigants for grain weevils by various investigators, who have reached the following conclusions: Carbon tetrachloride is ineffective under practiced conditions; hydrocyanic acid gas fails to kill weevils very far below the surface of the grain; sulphur dioxide has low toxicity, injures ironwork, destroys the germinating power of wheat, makes a sticky dough C4), and retards fermentation, the bread obtained being heavy and unfit for consumption; carbon dioxide is effective only in tightly sealed containers .and at relatively high concentrations; chloropicrin shows promise of being a practical fumigant, but is not yet com- mercially available; naphthalene is not very effective and has an objectionable odor; phosgene is poisonous to man, comparatively nonpoisonous to insects, and, because of its high vapor pressure, difficult to control; the toxicity of arsine to insects is comparatively low; the effect of cyanogen chloride as an insecticide is practically the same as that of hydrocyanic acid. It is evident, therefore, that there is great need for a fumigant which will be effective against injurious insects in wheat and other cereals and also noninflammable and nonexplosive and neither dangerous nor highly disagreeable to those who handle it. The object of the investigation here reported was to discover such a fumigant, which could be used in place of carbon disulphide. In connection with this investigation, much information on the relation of the chemical constitution of compounds to their toxicological action on insects was acquired. EXPERIMENTAL PROCEDURE The rice weevil ( Sitophilus oryza L.), the flour weevil ( Tribolium confusum Fab.), and the granary weevil (Sitophilus granarius L.) were used in most of the tests, and the Indian meal moth (Plodia inter- punctella Hbn.) was used in a great many. Adult insects of the weevils and the larvae of the Indian meal moth were used. The first series of experiments (Tables 1 and 2) were conducted in the apparatus described by Neifert (18). Four-liter glass jars, containing 10 to 20 live weevils of each species tested, were filled with a mixture of air of 40 per cent relative humidity and the vapor of the compound to be tested. After standing for 24 hours at room temperature (21° to 32° C.), the percentage of dead weevils was determined. (All specimens were examined after 24 hours, and also after 48 hours, to avoid reporting as dead those which might FUMIGATION AGAINST GRAIN WEEVILS 3 have been only stupefied.) To obtain the desired humidity, air was bubbled through bottles of sulphuric acid of the proper density, the vapor pressure of water in sulphuric acid-water mixtures being known for mixtures of various densities. In the less volatile com- pounds, the quantity of the compound present in the jar in the form of vapor was determined by passing a known volume of air over a weighed quantity of the compound. The loss in weight (in grams) of the compound divided by the quantity of air (in liters) gives the quantity (in grams) of substance present in the form of vapor per liter of air. The following factors were used in calculating the results reported : Grams per liter X 62. 43=pounds per 1,000 cubic feet. Grams per literX- 22 AX 100 — ___ =percentage concentration. gram-molecular weight Percentage concentration X gram-molecular weight X 0.02787 = pounds per 1,000 cubic feet. 1 cubic foot = 28.32 liters. 1 kilogram = 2.20462 pounds. Assuming the vapor of a compound to be a perfect gas, the mole- cular weight (in-grams) of this vaporized compound will occupy 22.4 liters under a pressure of 760 millimeters of mercury and at a tem- perature of 0° C. For example, the molecular weight of chloroform is approximately 119.4. Therefore, if 22.4 liters of space contains 119.4 grams of chloroform vapor (equivalent to 5.33 grams per liter), the percentage concentration is 100. Since the tests were made at about the same temperature on com- pounds differing greatly in vapor pressure, it follows that the more volatile compounds were tested in high molar concentrations and the slightly volatile compounds were tested in low molar concen- trations, which, however, were the maximum concentrations of the vapors possible at the temperature of the test. EFFECT OF VOLATILE ORGANIC COMPOUNDS ON WEEVILS The results of the fumigation of weevils in glass flasks at room temperature (21° to 32° C.) are shown in Table 1. Table 2 gives the formula, molecular weight, boiling point, and minimun lethal concentration for each compound tested. Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature C 21 ° to 32° C.) Test No. Tem- pera- ture Fumigant Concentration of fumigant Weevils killed after exposure for 24 hours Molar per- cent- age Pounds per 1,000 cu. ft. s. oryza s. gran - arius Tribo- lium Plodia Hydrocarbons: °C. P. ct. P. ct. P. ct. P. ct. 311 25.0 40. 31 78. 76 100 100 100 343 21. 5 14. 73 28. 78 0 0 0 347.. 22. 5 21. 26 41. 54 0 0 0 94 28. 5 i .40 1.43 0 0 0 10 83 23. 0 13. 00 30. 48 100 100 100 85 139 30. 5 14.00 32. 82 100 100 100 70 35 28. 5 13. 00 28. 29 100 100 100 100 130 28. 5 .90 1.96 0 0 0 0 133 29.0 1.60 3.48 10 30 80 0 138 30. 5 3. 80 8. 27 100 100 100 70 37 28.5 3.00 7. 70 75 50 100 80 23 30. 0 1.40 4. 14 60 0 0 14 24. 0 » .50 1. 79 76 25 25 and carbbn tetracblo- ride f75 per cent). 39 28.5 Anthracene .01 .05 0 0 0 0 1 Molecular weight of kerosene assumed to bo 128. 1 Concentration of naphthalene only. 4 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Table 1. — -Results of fumigation tests on weevils in glass flasks at room temperature ( 21 ° to 32° C.) — Continued Concentration Weevils killed after exposure of fumigant for 24 hours Test No. Tem- pera- ture Fumigant Molar per- cent- age Pounds per 1,000 cu. ft. ~s. oryza s. gran - arius Tribo- lium Plodia Bromides: °c. P. ct. P. ct. P. ct. P. ct. 50 . - 31.0 0. 94 6. 62 100 80 90 10 365 23.0 53. 50 162. 48 100 100 100 373 25. 5 16.80 51.02 100 100 100 376 25.0 14.20 43. 13 100 100 100 378 25. 5 8.80 26. 73 0 0 0 380 24.0 12. 70 38. 57 100 100 100 382 24.0 8.00 24. 30 100 100 100 404 24.0 7.60 23. 08 100 100 405 25.0 4.00 12. 15 100 100 410 27.0 4. 30 13.06 100 100 112 28.0 3. 30 10. 02 100 100 414 28.0 2. 90 8. 81 100 100 419 24. 0 1. 70 5. 16 0 0 420 24.0 1. 20 3. 64 100 0 421 24.0 .70 2. 13 0 0 366 23. 0 1. 30 6. 81 100 100 100 374 25. 5 .50 2. 62 100 100 100 377 25.0 .20 1. 05 100 30 80 379 25. 5 . 30 1. 57 0 0 80 381 24.0 . 30 1.57 75 50 100 383 24.0 .20 1.05 100 0 80 278 26.0 12. 79 43.84 100 100 100 280 26.0 13.84 47. 44 100 100 100 308 25.0 5. 50 18. 85 100 100 100 313 25. 0 7. 26 24. 89 100 100 100 326 24. 5 5. 54 18. 99 100 100 100 331 24.0 4.00 13.71 100 50 100 353 24. 0 4. 01 13. 75 0 0 0 367 24.0 4. 90 16. 80 100 100 100 369 24.0 4. 10 14. 05 100 100 100 385 24.0 2. 20 7. 54 0 0 0 483 26.0 15. 30 51. 59 100 100 486 26. 0 8. 00 26. 97 100 100 489 25.0 3.90 13. 15 100 100 275 26. 0 5. 27 20. 12 100 100 100 277 26. 0 5. 30 20. 24 100 25 100 307 25.0 2. 64 10. 08 100 100 100 312 25. 0 2. 56 9. 78 100 100 100 325 24. 5 1. 63 6. 22 80 25 100 330 24. 0 1.09 4. 16 0 25 15 54 . _ 32 0 . 77 3. 37 20 0 0 0 295 23. 0 . 56 2. 45 0 0 0 297 23. 5 . 55 2. 41 0 0 0 58 30.0 .12 .57 0 0 0 20 Chlorides: 518 25. 0 39. 50 93. 51 100 100 520 - - 25. 0 8. 10 19. 18 0 100 524 25. 0 16. 80 39. 77 100 100 525 25. 0 14. 20 33. 62 100 100 541 26.0 4. 40 10. 42 100 100 580 28. 0 3. 40 8. 05 0 0 581 28. 0 2. 20 5. 21 0 582 28.0 1. 20 2. 84 0 0 583 28. 0 1. 10 2. 60 0 19 31. 0 20. 00 66. 55 100 100 100 69 28. 5 2. 30 7. 65 0 0 0 0 79 30.0 15. 00 49. 91 100 100 100 100 104 28. 5 3. 20 10. 65 95 45 100 50 106 29. 0 7. 00 23. 29 100 100 100 80 109 29.0 13. 90 46. 25 100 100 100 100 11 30. 0 2. 00 8. 58 0 0 0 239 23.0 10. 05 43. 09 100 100 166 240 24.0 12. 21 52. 35 100 100 100 249 24.0 10. 31 44. 21 100 100 100 250 23.5 6. 86 29.41 100 100 100 89 . . 28. 5 20.00 55. 16 100 100 100 100 153 25. 5 10. 20 28. 13 100 100 100 100 156. 26.0 10. 90 30. 06 100 100 100 100 159 26. 5 2.00 5. 52 0 0 20 10 162 27.0 1. 30 3. 59 0 0 75 10 88 28. 5 4.00 .14. 87 100 100 100 100 114 28. 0 3. 80 14. 13 100 100 100 100 117 29.0 1.90 7. 06 100 90 100 65 120 30.0 .90 3. 35 95 0 0 65 123 30.0 .48 1.78 0 0 100 10 FUMIGATION AGAINST GKAIN WEEVILS 5 Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature {21° to 32° C .) — Continued Concentration Weevils killed after exposure of fumigant for 24 hours Test No. Tem- pera- ture Fumigant Molar per- cent- age Pounds per 1,000 eu. ft. s. oryza s. gran- arius Tribo- liurn. Plodia Chlorides— Con. °C. P. d. P. d. P. c/. P. d. 126 30.0 Trichloroetkane 0.22 0.82 0 0 70 0 497 24.0 2.80 10.41 100 100 499 26.0 1. 60 5.95 100 100 501 26.0 .80 2.97 100 100 503 24.0 .40 1.49 0 0 24.0 .20 .74 100 100 512, 513 23.0 .40 1. 49 0 0 516 24. 5 .20 .74 0 0 526 25.0 .80 2. 97 50 50 527. 25.0 .80 2. 97 100 75 528, 529 26.0 .50 1.86 50 50 530 25.0 1. 10 4. 09 100 100 531 25.5 1. 10 4. 09 100 100 66 - 28. 5 1.00 4. 68 60 0 0 50 78 29. 5 1. 10 5. 15 100 100 100 100 96 30.0 1. 10 5. 15 100 60 100 90 98 30. 0 .46 2. 15 60 50 100 90 100 30. 0 .23 1.08 70 50 100 60 87 28. 5 10.00 27. 58 100 100 100 100 113 28.0 12.00 33. 10 100 100 100 100 116 29. 0 6.00 16.55 100 100 100 100 119 29. 0 3. 00 8. 27 95 50 100 85 122 30. 0 1.40 3. 86 0 0 100 25 125 30. 0 .73 2. 01 90 20 100 20 84 31. 0 10.00 36. 62 100 100 100 100 95 30. 0 10.00 36. 62 100 100 100 100 97 30. 0 5.00 18.31 10 0 0 10 99 30. 0 2. 10 7.69 100 100 100 100 102 30. 0 1.05 3. 85 0 0 0 70 146 30. 0 5.00 18.31 65 60 100 20 150 30. 0 2.50 9. 16 50 15 80 0 574 26.0 Tetrachloroethylene 2. 60 12.02 100 100 100 575 26.0 1.30 6. 01 50 0 100 576 26.0 .60 2. 77 0 0 577. 26.0 .30 1.39 0 0 55— 32. 0 9. 02 24.37 100 100 100 80 141 29. 0 28.00 75. 65 100 100 100 100 145 30. 0 16. 80 45. 39 100 100 100 100 149 30. 0 6. 90 18. 64 60 0 90 0 536 26. 0 6. 10 19.21 100 100 537 26.0 2. 80 8. 82 100 100 538 26.0 1.20 3. 78 100 100 556 27.0 1. 10 3. 46 100 70 557 27.0 1.00 3. 15 100 80 558 27.0 .70 2. 20 100 60 559 27. 0 .70 2.20 100 0 92 28. 5 18.00 39. 40 0 0 65 50 1 26.0 1.00 3. 14 100 2 26.0 .25 .78 100 100 3 26.0 .10 .31 0 0 4... 26.0 .50 1. 57 0 80 100 8 30. 0 2. 00 6. 27 100 100 100 9 30.0 1.00 3.14 100 50 100 15 29. 5 2.00 6. 27 75 32 20 16 30. 5 2.00 6. 27 90 50 20 28. 5 7.70 24. 15 100 100 100 95 138 30. 5 .60 2. 46 0 o 0 0 6... 26.0 1.00 4. 10 50 50 10 30.0 1.00 4. 10 85 75 100 7 30.0 2.00 8. 19 100 100 13 30.0 1.00 4. 10 100 50 100 Fluorides: 465.... 24.0 14.30 38. 29 100 100 467.. 24.0 7.20 19. 28 100 100 469 24.0 3.40 9. 10 100 100 471.. . 24. 5 do 1.90 5. 09 100 0 474 23. 0 .do .. 2. 20 5.89 0 100 476 23. 0 .do 1.90 6.09 100 25 60 30.5 Dilluorodiphenyl .002 .01 0 0 0 10 Iodides: 302 25. 0 n-Butyl iodide .. 1. 61 8. 20 100 100 100 303 25. 0 1. 54 7. 90 100 100 100 316 24. 5 do .79 4. 05 100 100 100 337 21. 5 . .do. .43 2. 21 0 0 0 340 24.0 do .36 1.85 0 0 0 6 BULLETIN 1313, U. S. DEPARTMENT OE AGRICULTURE Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature {21° to 32° C.) — Continued Concentration Weevils killed after exposure o f fumigant for 24 hours Test No. Tem- pera- ture Fumigant Molar per- cent- age Pounds per 1,000 cu. ft. s. oryza s. gran- anus Tribo- lium Plodia Alcohols and phe- nols: °c. P. ct. P. ct. P. ct. P. ct. 34 28. 5 Methyl alcohol .. 16.00 14.28 100 100 100 100 129. 28. 5 do_ 1.00 .89 0 0 o o 132 29.0 do ... 2. 00 1. 79 0 0 0 0 135 30.0 do. 4. 50 4.02 30 15 5 10 64 30. 5 Ethyl alcohol.. . 8.00 10. 27 .0 0 0 15 21... 30. 5 7i-Propyl alcohol.. 3. 50 5. 86 100 100 100 71.. 32. 0 do _ 1. 70 2. 85 96 50 100 100 Ill 29. 0 do. _ 1. 70 2. 85 100 100 100 100 137... 30. 5 ... do ... . 1. 50 2. 51 20 0 15 85 81 30. 0 1.70 3. 51 100 60 100 100 20 31. 0 Isoamyl alcohol. .80 1. 96 100 100 80 142 29. 0 do .77 1.89 80 60 100 100 144 30. 0 __do .33 .81 0 0 0 0 148. 30. 0 __ __do .20 .49 0 0 0 0 68 28. 5 Geraniol . . 10 .43 10 0 0 0 59. 30. 5 Menthol.. .08 .35 0 0 0 80 46 30.0 Thymol .09 .38 0 0 0 0 Aldehydes: 67 28.5 Chloral hydrate .20 .92 20 0 0 0 289. 25. 0 m-Butyraldehyde 11. 15 22.40 100 100 100 291 25. 0 do 9. 46 19. 00 100 100 100 321 26.0 . _— do ... 5. 53 11. 11 100 100 100 335. 21. 5 _ __do 3. 27 6. 57 0 0 100 338 24.0 do. . 2.59 5. 20 0 0 90 464. . . 24. 0 Crotonaldehyde 4. 00 7. 81 100 100 466 24.0 _do 2.00 3.91 100 100 468 24.0 do. 1. 30 2.54 100 100 470 24. 5 do _ __ .50 .98 100 0 473 . .. 23. 0 . __.do .60 1. 17 100 100 475. 23.0 _ _ _do_ _ .50 .98 100 75 18 30. 5 Furfural .80 2. 14 100 100 20 52.. . 31. 0 Benz aldehyde .27 .80 100 100 55 75 Ketones: 33 28.5 Acetone 15.00 24. 27 100 100 100 100 128 28. 5 __ __do 2. 00 3.24 5 10 10 0 131. . 28. 5 _ ___ do 4. 10 6. 63 95 100 15 25 134. 30. 0 _ .do 8.20 13.27 100 100 100 100 219. 24.0 Chloroacetone. 1. 53 3. 94 100 100 100 220. 23. 0 _ ...do .... .70 1. 80 100 100 100 232. 24. 0 do .42 1. 08 100 100 100 243.. . 26. 0 do .22 .57 0 0 0 268 25.0 do .42 1. 08 100 100 100 270. . 25. 0 do. ... .26 .67 100 100 100 80. . 23. 0 Ethyl methyl ketone . 15. 00 30. 13 100 100 100 100 140 29. 0 do __ _ 13. 30 26. 72 100 100 100 100 143 29.0 do. ... 7. 20 14. 46 100 100 100 100 147 30.0 do __ ... 3. 40 6. 83 100 70 100 80 151 30.0 . ..do 1. 80 3. 62 100 80 100 0 304 25.0 Mesityl oxide. 1. 53 4. 18 100 100 100 309 25.0 do _. ._ 1.42 3.88 100 100 100 341. . 21.0 __ ..do .39 1.07 0 0 0 345. 22. 5 do .55 1.50 100 50 0 352. 23. 0 do .60 1.64 100 100 0 53 31.0 Acetophenone... . .16 .54 100 45 0 50 6 30.0 ] Chloroacetophenone (5 per f 3.10 .43 100 100 12 30.0 1 cent) in carbon tetra- i 3.01 .04 0 0 0 63 30.5 J chloride (95 per cent). l 3.17 .73 0 0 0 5 Esters: 217. 26.0 Methyl formate. . . 56.00 93.71 100 100 100 222. . 23. 0 do ... 20.37 34.09 100 100 100 234. 24. 5 do 11. 62 19.44 100 100 100 241 25.5 do. _ 6. 35 10.63 100 100 100 259. 26. 0 do ... 4.54 7. 60 100 100 100 260. 26. 0 __ __do 3. 64 6.09 100 100 100 271. 24. 5 do ... 3. 64 6. 09 0 0 0 272 24.5 _ _ .do 4. 10 6. 86 100 100 100 273 . 24. 5 do 2. 65 4. 43 100 100 100 474A, 475A 24. 5 do 3.50 5. 86 100 100 476A 24. 5 do - 2.00 3. 35 100 100 506. 24.0 do 1.30 2. 18 100 100 507 ... 24. 0 do. .. . .90 1. 51 100 100 284. . 24. 5 20. 37 42.05 100 100 100 286 24. 5 do 22.41 46. 26 100 100 100 319 . 26. 0 15.04 31.04 100 100 100 328 24.0 do 10. 00 20.64 100 100 100 3 Concentration of chloroacetophenone only. FUMIGATION AGAINST GRAIN WEEVILS 7 Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature ( 21 0 to 32° C .) — Continued Test No. Tem- pera- ture Fumigant Concentration of fumigant Weevils killed after exposure for 24 hours Molar per- cent- age Pounds per 1,000 eu. ft. s. oryza s. gran - arius Tribo- lium Plodia Esters— Con. °C. P. ct. P. ct. P. ct. P.ct. 333 24.0 Ethyl formate 6.41 13. 23 100 100 100 357 - 24.0 4. 89 10. 09 100 100 100 359 23.0 _ ___do 3. 71 7. 66 100 100 100 389 25.0 do 1. 80 3. 72 100 100 100 391 25.0 do .60 1.24 100 100 100 406 25.0 do 1. 30 2. 68 100 100 407... 25.0 _ ___do 1. 10 2. 27 100 80 408. : 25.0 do .90 1.86 100 100 411A 23.0 do 1. 10 2. 27 100 100 412A, 422 23.0 do .80 1. 65 100 100 423 — 23.0 do .40 .83 0 0 424 23.0 do.__ .60 1. 24 100 0 426,428 25.0 do .50 1. 03 0 o 484 26.0 Isopropyl formate 15. 50 38. 05 100 100 487 26.0 Idol. 8.20 20. 13 100 100 490 25.0 do__ 3.90 9. 57 100 100 492. 25. 5 2.00 4.91 100 100 494 25. 5 do__ 1.00 2. 45 50 0 514 24.0 do 1.20 2. 95 100 100 515 25.0 do .90 2. 21 100 100 532 26. 5 do___ 1. 10 2.70 100 100 533 26. 5 .90 2. 21 100 100 542 26. 0 .80 1.96 100 80 543 26.0 do . 50 1. 23 100 0 166 27.0 Methyl cyano formate .20 .47 16 0 100 100 167 27. 5 .32 .76 88 90 100 100 169 28.0 .08 .19 24 55 15 95 175 27. 0 .90 2. 13 100 100 100 100 176 27. 5 .68 1.61 100 100 100 100 177 27. 5 .40 .95 95 90 100 100 178 27. 5 .30 .71 50 50 100 100 38 27. 5 3. 50 9. 96 100 100 100 100 73 32. 0 2. 50 7. 11 100 100 100 100 155 25. 5 3. 70 10. 53 95 90 100 100 158 26.0 1.90 5.41 90 50 100 100 161 27.0 1.00 2. 85 0 0 0 0 44 30. 0 .76 2. 46 70 20 95 100 282 26.0 1.68 5.44 0 0 0 283 26. 0 1. 56 5. 05 0 0 0 45 . 30.0 1. 50 5. 44 100 100 100 100 75 31. 0 .90 3. 26 100 100 100 100 101 30.0 .54 1.96 20 50 75 85 103 30. 0 .24 .87 0 0 0 0 294 23. 0 .77 2. 79 0 0 0 296 23. 5 .66 2. 39 0 o 0 Ethers: 292 25. 0 4.86 16.00 100 100 100 293 25. 0 4. 34 14. 29 100 100 100 322 26. 5 2.27 7. 47 50 20 100 344 21. 5 1.72 5. 66 0 0 80 348 22. 5 1.70 5.60 100 100 100 351 23. 0 1.64 5. 40 50 20 100 621 27. 0 4. 40 14. 49 100 100 523 27. 0 1. 10 3. 62 100 0 274 26.0 2. 34 5. 25 100 100 100 275 26. 0 2. 22 4. 98 100 100 100 218 26. 0 4. 87 15. 60 100 100 100 221 23. 0 3. 89 12.46 100 100 100 233 24. 0 1. 13 3. 62 100 100 100 244. 26.0 .62 1.99 50 50 90 267.. 25.0 1.21 3.88 100 100 100 269 25.0 .70 2. 24 0 0 0 Chlorohydeins: 517 25.0 Ethylene chlorohydrin 1.40 3. 14 0 0 519 25. 0 .30 .67 0 0 288 24. 5 1.95 5. 03 100 100 100 290 25. 0 1.93 4. 98 100 100 100 320 20. 0 1. 19 3.07 100 100 100 329 24.0 .73 1.88 100 100 100 334 24. 0 .59 1.52 100 100 100 358... 23. 0 .42 1.08 100 100 100 360 23. 0 .41 1.06 100 100 100 363 23. 0 .25 .64 100 100 100 388 25. 0 .20 . 52 100 100 100 390 25. 0 . 10 .26 100 100 100 397 25. 0 .09 .23 100 50 398 25. 0 do .09 .23 100 0 8 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature {21° to 82° C .) — Continued Concentration of fumigant Test No. Tem- pera- ture Fumigant Molar per- cent- age Pounds per 1,000 cu. ft. Sulphur com- pounds: °c. 115 28. 0 Carbon disulphide 46. 00 97. 59 118. 29. 0 24. 70 52. 40 121 30. 0 12. 30 26. 10 124 30. 0 6. 40 13. 58 127 30. 0 3. 00 6. 36 164 27.0 3. 30 7.00 165 27. 0 2. 50 5. 30 168 27. 5 do 1. 10 2. 33 170- 26. 0 1. 00 2. 12 171 26. 0 . 50 1. 06 172 27. 0 .47 1. 00 173 27. 0 . 50 1. 06 174 27. 0 . 11 . 23 93 - - 28. 5 19. 00 32. 89 29 31. 5 n-Butyl mercaptan 3. 00 7. 54 72.- 32. 0 do".— 1 2. 00 5. 03 152 25. 5 5. 40 13. 57 154 . 25. 5 2. 30 5. 78 157 26.0 .90 2. 26 160. 26. 5 . 30 . 75 163 27.0 . 10 .25 216 26. 0 Methyl sulphide 53. 80 93. 14 223 23. 0 20. 98 36. 32 235 25. 0 9. 45 16. 36 242 25. 0 10.44 18. 07 256 23. 5 9. 75 16.88 257 24. 0 9. 49 16. 43 258 23. 5 5. 00 8. 66 265 25. 0 12.87 22. 28 266 25. 0 11. 54 19. 98 215 26. 0 4. 91 12. 34 224 24. 0 2. 92 7. 34 236 25. 0 1. 50 3. 77 245 25. 5 do 1. 72 4.32 261. . 24. 5 3. 65 9. 17 262 25.0 3. 72 9. 35 264 25. 0 2. 34 5. 88 213 26. 6 1.84 3. 75 226 24. 0 . 78 1. 59 237 25. 0 .36 .73 246 26. 0 . 20 . 41 253. 26. 5 . 20 .41 254 26. 5 . 14 .29 255 26. 0 .09 . 18 263 . 26. 0 . 10 . 20 214 26. 6 Ethyl thiocyanate . 1.04 2. 53 225 24. 0 .46 1. 12 229 23. 0 .46 1. 12 231 24. 0 . 19 .46 238 23. 0 .71 1.72 402 24. 0 .80 1.94 51 31. 0 1. 14 3. 15 77 29. 5 .70 1.93 107 29.0 .34 .94 110 29.0 . 62 1. 71 30 28. 5 . 10 . 23 31 . 28. 5 . 16 1. 16 28 30.5 Perchloromethyl mer- captol. 1.00 5. 18 90 28.5 p-Toluenesulphochloride .. .15 .80 Nitriles and iso- nitriles: 305 25. 0 11. 60 13. 27 310 25.0 do 12. 13 13. 87 342 21.0 3.47 3. 97 346 22. 5 _ __ _do 4. 97 5. 68 371 26. 5 3. 90 4.46 372 26. 5 -.-do . 5. 70 6. 52 430! 25. 0 1.30 1. 99 414A 24. 0 n-Valeronitrile 1. 10 2. 55 416 24.0 do .60 1. 39 418 24.0 .30 .69 26 30. 0 Phenylisonitrile.-. _ .. .12 .34 425 25. 0 Phenylacetonitrile . . .10 .33 427 25. 0 _ _do .03 . 10 429 25. 0 do __ .02 .07 437. 25. 0 .07 .23 438 25.0 do .03 . 10 Weevils killed after exposure for 24 hours s. oryza s. gran- arius Tribo- lium Plodia P. ct. P. ct. P. ct. P. ct. 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 35 70 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 20 80 0 0 0 0 10 0 0 0 0 100 100 100 100 100 100 50 0 100 100 100 100 90 0 100 100 100 100 0 0 50 60 0 100 100 100 100 100 100 100 100 100 100 20 10 90 95 80 80 100 100 100 100 100 100 50 0 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 0 0 0 100 20 30 100 100 100 100 100 50 5 0 100 100 100 50 0 0 0 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 0 0 0 0 0 0 0 0 60 100 25 0 0 0 0 100 100 100 100 100 100 100 0 0 0 100 100 100 50 0 0 100 100 100 0 0 100 100 50 30 0 0 0 0 0 0 100 90 0 0 0 0 0 0 0 0 FUMIGATION- AGAINST GRAIN WEEVILS 9 Table 1. — Results of fumigation tests on weevils in glass flasks at room temperature {21° to 82° C.) — Continued Test No. Tem- pera- ture Fumigant Concentration of fumigant Weevils killed after exposure for 24 hours Molar per- cent- age Pounds per 1,000 cu. ft. s. oryza s. gran- arius Tribo- lium Plodia Nitrites: °C. P. ct. P. ct. P. ct. P. ct. 413A 24. 0 9. 10 26. 15 100 100 415A 24. 0 4. 30 12. 36 90 90 417 24. 0 do 2. 20 6. 32 100 100 434 25. 0 1. 60 4. 60 100 100 435 25. 0 1. 10 3. 16 100 100 436 25. 0 do 1. 00 2. 87 100 100 442 24. 0 do .60 1. 72 100 100 443 24. 0 do_ .60 1. 72 100 66 444. 24. 0 __do .30 .86 50 8 279 26. 0 Isoamyl nitrite 4. 70 15. 34 100 100 100 281 26. 0 ___ _do_ 4. 84 15. 80 100 100 100 318 24. 5 do 2. 36 7. 70 100 100 100 327 24. 0 1.46 4. 77 100 100 100 332 24. 0 do 1. 02 3. 33 100 50 100 354 24. 0 do 1. 13 3. 69 0 0 0 368 24. 0 1. 40 4. 57 100 100 100 370 26. 0 do 1. 10 3. 59 100 75 100 384 24. 0 do .80 2. 61 100 100 100 399 25. 0 do 1. 20 3. 92 100 100 400 25. 0 do .80 2. 61 100 100 411 23.0 do .70 2. 28 100 50 413 24. 0 do .60 1. 96 100 100 415 24. 0 do .80 2. 61 100 100 NlTRO COMPOUNDS: 298 23. 5 4. 35 7. 40 100 100 100 300 25. 0 4. 59 7.81 100 100 100 323 26. 5 do 3. 57 6. 07 100 100 100 336 21. 5 do 1. 27 2. 16 0 0 0 339. 24. 0 1.03 1. 75 0 o 0 32 28. 5 . 10 .34 0 0 0 0 49 31. 0 . 12 .41 0 o 0 0 82 30. 0 . 12 .53 0 o 0 0 85 31. 0 .28 1.31 0 o 0 0 86 28. 5 a -Nitronaphthalene . 10 .48 0 o 0 0 Amines: 496 24.0 20. 60 41.98 100 100 498. . 24. 0 11. 80 24. 04 100 100 500 26.0 do 5. 80 11. 82 100 100 502 25. 0 do 2.60 5. 30 100 100 504 23.5 do 1. 10 2. 24 100 100 509 24. 0 do .90 1.83 100 100 511 23. 0 __do .50 1.02 100 100 534 26. 0 do .80 1. 63 50 50 535 26.0 . 50 1. 02 50 100 482 26. 0 n-Butylamine 10. 20 20. 78 100 100 485 26. 0 5. 30 10. 80 100 100 488 27. 0 2.80 5. 71 100 100 491 25. 5 1. 30 2. 65 100 100 493 25. 5 .60 1. 22 100 100 508 24. 0 .30 .61 50 0 510 24. 0 .20 .41 0 0 43 29. 5 1. 89 4. 90 0 0 0 0 47 29. 0 .13 .39 0 0 0 0 66 30. 0 . 15 .51 0 0 0 0 67 30. 0 Ethylbenzylaniline .03 . 18 0 0 0 0 62 30. 5 Acetphenylenediamine .07 .29 0 0 0 0 91.. 28. 5 .016 .06 0 0 0 0 Miscellaneous ni- TROGENOUSCOM- pounds: 48 30. 0 1.69 3. 72 100 100 100 100 76 . 31. 0 .84 1. 85 100 100 100 100 105 29. 0 . 17 .37 0 0 0 0 108 29. 0 .39 .86 10 5 10 20 112... 29. 0 .83 1. 83 100 100 76 65 61 '. 30.6 Hexamethylenetetramine. . .002 .008 0 0 0 0 Inorganic com- pounds: 477A . . . 23. 5 100.00 122. 64 100 100 478A . . . 24. 5 100. 00 122. 64 66 40 479.. 24. 5 50. 00 61. 32 100 25 480. 24. 5 33. 00 40. 47 0 0 481. 24. 5 60. 00 61. 32 100 100 495. . 25. 5 60. 00 01.32 100 0 622. . 27. 0 60. 00 61.32 100 0 41... 29. 0 6. 00 16. .58 100 100 100 100 74 32.0 do 4. 50 14.92 100 100 100 100 15685°— 25f 2 10 BULLETIN 1313, U. S. DEPARTMENT OP AGRICULTURE Table 2. — Formulae, molecular weights, boiling points, and lethal concentrations ( from results in Table 1 ) of fumigants Lethal concentration 3 Fumigant Hydrocarbons: Amylene.. Kerosene Cyclohexane Benzene Toluene.. o-Xylene Naphthalene Anthracene... Bromides: Bromoform Ethyl bromide Ethylene bromide. n- Propyl bromide Allyl bromide 7i-Butyl bromide. Bromobenzene Benzyl bromide Chlorides: Methylene chloride Chloroform Carbon tetrachloride ... Acetylene dichloride Ethylene chloride Ethylidene chloride Trichloroethane Trichloroethylene Tetrachloroethane Tetrachloroethylene Propylene dichloride ... Isopropyl chloride M onochlorobenzene o-Dichlorobenzene p-Dichlorobenzene o and p- Dichlorobenzene Fluorides: Fluorobenzene Difluorodiphenyl Iodides: «-Butyl iodide.. Alcohols and phenols: Methyl alcohol Ethyl alcohol n-Propyl alcohol... w-Butyl alcohol Isoamyl alcohol Geraniol Thymol Menthol Aldehydes: Chloral hydrate n-Butyraldehyde Crotonaldehyde Furfural.. Benzaldehyde Ketones: Acetone Chloroacetone Ethyl methyl ketone... Mesityl oxide Acetophenone.. Chloroacetophenone Esters: Methyl formate Ethyl formate. Isopropyl formate Methyl cyano formate.. m-Propyl acetate n-Butyl acetate Isobutyl acetate Isoamyl acetate Ethyl-n- valerate Formula Mo- lecular weight 1 Boiling point 2 3 (molar percentage) s. oryza s. gran- anus Tri- bol- ium Plo- dia °C. 70. 105 22-37 40.3 40.3 40.3 150-280 84. 126 80.8 4 13. 0 4 13.0 4 13.0 78. 078 79-81 3. 8 3. 8 3. 8 13.0 C8H6CH3 92. 099 109-110. . 3.0 C6H4(CH3)2 106. 120 143.5-144.5 128. 114 218 178. 150 351 CHBrs. 252. 773 142-151 ... .9 108. 970 38-40 2.9 4 8.0 2.9 187. 882 127-132 .5 .5 . 5 122. 991 70-71 4. 1 4. 1 4. 1 120. 975 69-71 4 3. 9 4 3. 9 137. 012 100-101 . 2. 6 2. 6 1. 6 156. 990 154-155 . 171.011 198-199 84. 941 40.5-42. . 4. 4 4.4 CHCls 119. 393 58-61.5 7.0 7.0 3.2 13.9 CCh .. 153. 845 76-77 6. 9 6. 9 6. 9 96. 946 55 9.0 9. 0 9. 0 16.8 98. 962 83-84 6.0 6. 0 4 . 7 6.0 C2H4CL 98. 962 59.5-61.5 . 10. 2 10. 2 10. 2 10. 2 133. 414 74.5. . 8 3. 8 1. 1 3.8 C2HC13 131.398 85.5-87 10. 0 10.0 10.0 10.0 C2H2C14 167. 866 144-146.... 1. 1 1. 1 1. 1 1. 1 C2C14 — - 165. 850 119-121 2. 6 2. 6 1. 3 112.983 95-96 4 . 7 1. 2 (CH3)2CHC1 78. 531 34-36 CeHsCl 112. 530 128-132 2. 0 2. 0 2.0 C6H4Cl2 - 146. 982 178-181 ... 146. 982 172-174 146. 982 1.0 2. 0 1.0 96. 070 84.9 4 1.9 2.2 190. 124 254-255 C4HeI 184. 012 129-131 .8 .8 .8 CHjOH.... 32. 037 65 16.0 16.0 16.0 16.0 46. 058 78 60. 079 96-98. 1. 7 3. 5 1. 7 1.7 GiHdOH. 74. 100 114-118 41. 7 4 1.7 4 1.7 C5H11OH 88. 121 130-132.... .8 .8 .8 .8 154. 194 230 (CH3)2CHCeH3(CH3)OH . 150. 162 228-232 156. 210 210 CC13CH(0H)2 165. 414 97. 5. C3H7CHO 72. 084 75-77.5 5. 5 5. 5 3.3 CHjCH-CHCHO 70.068 104-105 4. 5 .6 C(H3OCHO 96.057 161 . i. 8 4. 8 106. 083 179 4. 3 4. 3 CHaCOCHs.. 58.063 55-56 8.2 4. 1 8. 2 8.2 CH2C1C0CH3 92. 515 117-119 .26 .26 .26 72. 084 79-80 4 1.8 7.2 4 1. 8 7. 2 (CH3)2C-CHCOCHj 98. 110 130-131 .55 .6 1.4 120. 104 202. 4. 16 CcHsCOOHjCl 154. 556 247 .10 . 10 HCOOCHs 60. 042 31.5-32.5... 4. 9 4 2. 65 4. 9 74.063 53.5-55 .6 4. 6 .8 HCOOCH(CH3)2— 88. 084 63-64 4. 5 .9 CNCOOCH3 85. 047 100-101.... .7 .7 .2 .2 102. 105 99-102 2. 5 2. 5 1. 9 1.9 CH3COOC4H9 116. 128 124-126 4. 76 CH3COOC4H9- 116. 126 115-117 CHsCOOCsHn 130. 147 138-140.... .9 .9 .9 .9 C4H9COOC2H5 130. 147 144-145.5.. 1 Based on international atomic weight values for 1922. 5 The boiling points given are those of the substances as tested, not those of these compounds of the highest degree of purity. 3 Thel ethal concentration here reported represents the minimum percentage concentration which con- sistently caused 100 per cent mortality after exposure for 24 hours. Since in some cases the concentra- tions tested varied decidedly and because of some variation in the experimental results, the results here given do not necessarily represent the exact minimum lethal concentration. * Minimum concentration tested. FUMIGATION AGAINST GRAIN WEEVILS 11 Table 2. — Formulae, molecular weights, boiling 'points, and lethal concentrations ( from results in Table 1 ) of fumigants — Continued Mo- Boiling point Lethal concentration (molar percentage) Fumigant Formula lecular weight S. oryza s. gran - arius Tri- bol- ium Plo- dia Ethers: 118. 142 °C. 102-104 4.3 4.3 1.7 Chloromethyl ether... s-D ichlorom ethyl ether. Chloroh ydrins : Ethyleneehlorohydrin CH2CIOCH3 80. 510 65-60 4 2. 2 4 2. 2 4 2. 2 CH3CIOCH2CI 114. 962 100-108- - 1.1 1.1 1.1 CH2CICH2OH 80. 510 126-127 C3H6C10... 92. 515 115-117 4. 09 4.io .10 Sulphur compounds: Carbon disulphide CS2 76. 125 46 1.0 1. 1 1. 1 1.0 C2H5SH 62. 118 34.5-35.5— 4 19. 0 4 19. 0 4 19. 0 4 19.0 ri-Butyl mercaptan.. . C4H9SH 90. 160 96-98 .9 2.0 2.0 2.0 (CH3)3S 62. 118 38-39 10.4 10.4 9.5 90. 160 92-93 2. 9 2.9 2.3 Methyl thiocyanate .. Ethyl thiocyanate Allyl isothiocyanate . . Cyanogen sulphide... CH3SCN 73. 102 130-131 . 1 . 14 . 14 C2H5SCN— 87. 123 146-147 .8 1.0 .8 CsHsNCS- 99. 128 143-151 4. 3 4. 3 4. 3 4. 3 (CN)3S 84. 086 (Sublimes 30-40°) 260. 323 Perchloromethyl mer- captol. p-Toluenesulphochlo- ride. Nitriles and isoni- triles: CCBSCli 185. 905 146.5-148— 4 1.0 190. 611 145-146 (15 mm.) 76-82 4. 15 ch3cn. 41. 042 5.0 6.0 5.0 C2H6CN— 55. 063 96-97. C4H9CN- 83. 105 139-141 1. 1 1. 1 CeHsNC 103. 083 165-166 Phenylacetonitrile Nitrites: C0H5CH2CN 117. 104 231. 7 .1 103. 100 77-79. .6 1.0 117. 121 96-99 4. 6 1.4 .8 Nitro compounds: 61. 037 98-101 3.6 3.6 3.6 C0H5NO2 123. 078 209.4 p-Chloronitrobenzene ClCflH4N02 157. 530 242 CBH4(N02)2 - 168. 078 297 o'-Nitronaphthalene . . Amines: 173. 114 304 (C2H5)3NH 73. 116 55-56 .9 4. 5 73. 116 76-78 .6 .6 93. 094 182 Methylaniline CiH[NHCH3 107. 115 193.5 C6H6N(CH3)2.-- 121. 136 192 Ethylbenzylaniline. . . Aeetphenylenediamine a-Naphthylamine Miscellaneous nitrog- enous compounds: CeHsN (C2H5) CH2C0HS 211. 219 285-286— NH3CtH4NHC0CH3 150. 136 143. 130 300 CsHjN 79. 073 116-118— .83 .83 .83 .83 Hexamethylenetetra- mine. Inorganic compounds: 140. 158 CO2 44. 005 —78.5 50.0 100.0 SOCI2 . . 118.980 78 4 4. 5 4 4. 5 4 4. 5 4 4. 5 4 Minimum concentration tested. * HYDROCARBONS Against grain weevils, a 100 per cent kill was obtained with amy- lene at a concentration of 40.31 per cent, with cyclohexane at a con- centration of 13 per cent, and with benzene at a concentration of 3.8 per cent. Against all of the insects except Tribolium, kerosene, tolu- ene, o-xylene, naphthalene, and anthracene were not more than 75 per cent effective. Trillat and Legendre (26) found that benzene and toluene vapors at a concentration of 10 grams per cubic meter were insufficient to kill mosquitoes after exposure for an hour at 20° to 28° C. Holt (.9) tried the effect of benzene, toluene, xylene, naphthalene, anthracene, “ benzoline,” heptane, petrol, petroleum ether, and par- affin oil (boiling point, 150° to 289° C.), in concentrations of 1 drachm 12 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE to 1,250 cubic centimeters, on roaches ( Periplaneta orientalis L.) con- fined in a glass-stoppered 1,250-cubic centimeter bottle. Benzene and toluene showed the same toxicity. In general, the higher the boiling point the longer the time necessary to kill. “ Benzoline” (boiling point, 60° C.), which killed in 17 minutes, was the most toxic. McClintock, Hamilton, and Lowe (13) compared the toxicity of thq vapors of naphthalene, kerosene, Australene, Oregon fir turpen- tine, oil turpentine, Michigan wood turpentine, and oil of Pinus pdlustris on bedbugs, cockroaches, house flies, clothes moths, and mos- quitoes in an 800-liter hood. Naphthalene was the most toxic. Jewson and Tattersfield (10) found that naphthalene vapor had no apparent effect on mites (Aleurobius farinx De G.), even after exposure for 16 hours. Lloyd (12) found pure naphthalene, free from tarry acids, to be a poor fumigant against adult greenhouse white flies (Asterochiton vaporariorum Westw.) . At a temperature of 69° to 72° F., 0.5 gram of pure naphthalene killed only 3 per cent of the adult white flies con- fined in a half-gallon glass-stoppered jar during exposure for an hour. Lefroy (11) found that xylene, turpentine, cymene, and pseudocu- mene killed 100 per cent, and that eucalyptus oil killed some meal- worms dipped in the liquid. Russell (22) found toluene more effective than carbon disulphide in the partial sterilization of sick soil in which tomatoes were growing. Titschack (25) found the vapor of benzene to be very much less effective than carbon disulphide against the eggs, larvae, and moths of the clothes moth ( Tineota biselliella Hum.) . Xylene also showed a low toxicity, but naphthalene was effective. Richardson and Smith (21) found the toxicity toward the black aphis (Aphis rumicis L.) to increase from benzene through toluene to xylene, but in all cases the tolerance of the host plant, nasturtium, was much less than the minimum toxic concentration. Cyclohexane was twice as toxic as benzene. Tattersfield and Roberts (24) found that anthracene and phenan- threne were nontoxic to wireworms; that mesitylene, p-cymene, and naphthalene had marginal toxicity; and that n-pentane, 7i-hexane, 71-heptane, benzene, toluene, and m- and p-xylene had low toxicity. Of the hydrocarbons examined only pseudocumene had moderate toxicity. Moore (15) found that the toxicity of petroleum ether, benzene, toluene, xylene, gasoline, camphene, naphthalene, and kero- sene toward flies increased with diminishing volatility. The low toxicity and ready inflammability of all the hydrocarbons tested make this class of organic compounds unpromising in the search for a practical fumigant. BROMIDES The most effective bromide tested during the investigation re- ported in this bulletin was ethylene bromide, which killed 100 per cent of the weevils at a concentration of only 0.5 per cent. The order of toxicity of the other bromides follows : Bromoform, 77-butyl, ethyl, allyl, 7i-propyl, and benzyl bromide, and bromobenzene. Bromoform and monobromobenzene were found by Tattersfield and Roberts (24) to be moderately toxic to wireworms. Moore (14) tested the action on flies of bromobenzene, p-dibromobenzene, o- and p-bromotoluene, and bromoxylene. The disubstitution products were more toxic than the monosubstitution products, and the oromine FUMIGATION AGAINST GRAIN WEEVILS 13 compounds were more toxic than the corresponding chlorine com- pounds. From tests on potato-beetle eggs with bromoform, tri- methylene bromide, o-bromotoluene, and bromoxylene, Moore and Graham (17) concluded that the toxicity of organic compounds to insect eggs increased with decreasing volatility. Moore (15) also tested the action of bromoform, brometone (tribromo tertiary butyl alcohol) , and ethylene bromide on flies. Bertrand and Rosenblatt (2) found benzyl bromide to be more toxic than carbon disulphide, but less toxic than monochloroacetone to the larvae of (Bornbyx) Malacosoma neustria L. Holt (9) found that it took nine minutes for the vapor of bromo- same concentration. Organic bromides have the disadvantage of being costly. While the bromides are in general more effective than the corresponding chlorides, they can not be expected to give an economical fumigant, because liquid bromine is usually quoted at a price four to five times that of chlorine; furthermore, 35.5 units by weight of chlorine are the chemical equivalent of about 80 units by weight of bromine. Ethylene bromide and ethyl bromide (Tables 3, 4, 5, 6, 7, 8) are the only bromides which seem worth a further trial. Trichloroethane, s-tetrachloroethane, propylene dichloride, and a mixture of o- and p-di chlorobenzene gave 100 per cent mortality in concentrations of 2 per cent or less. Other chlorides showed the following order of toxicity: Monochlorobenzene, p-dichlorobenzene, tetrachloroethylene, methylene . chloride, ethylene chloride, carbon tetrachloride, chloroform, acetylene dichloride, trichloroethylene, ethylidene chloride, and isopropyl chloride. Tattersfield and Roberts (24) experimented with wireworms. Dichloroethylene, trichloroethylene, carbon tetrachloride, chloro- form, and tetrachloroethane, in the aliphatic series, and mono- chlorobenzene and o-chloro toluene, in the aromatic series, had a low toxicity; 1, 2, 4-trichlorobenzene, monochloroxylene, p-dichloroben- zene, and benzotrichloride had a marginal toxicity; o-dichloroben- zene and benzal chloride had a moderate toxicity; and benzyl chloride had a high toxicity. Parker and Long (19), using several chlorides against the larvae of Trogoderma Tchapra, Arrow, at a concentration of 10 ounces per 1,000 cubic feet and an exposure of 1,000 minutes, obtained a mortality of 11 per cent with carbon tetrachloride, 27.7 per cent with tri- chloroethylene, 73.3 per cent with tetrachloroethane, and 77.7 per cent with pentachloroethane. Holt (9), Trillat and Legendre (26), Bertrand and Rosenblatt (2), Lefroy (11), McClintock, Hamilton, and Lowe (13), Titschack (25), and many other investigators report that chloroform has a low toxicity against various insects. Altson (1), using the vacuum method commonly employed with hydrocyanic acid gas, found tetrachloroethane useless for the fumi- gation of beetles in wood. He recommends both o- and p-dichloro- benzene as a deterrent to timber beetles. Lloyd (12) reports that tetrachloroethane gives good results against the white fly in greenhouse fumigation, but is without CHLORIDES 14 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE effect on the eggs. About half a pint per 1,000 cubic feet should be used, and the fumigation should be continued for not less than 18 hours. Lloyd states that: It has been used for a wide variety of plants, including tomatoes, and no damage has resulted except in one case when the foliage of three young sycamores {Acer pseudo-planatus) 4 growing in pots turned brown the day after the fumi- gation and was subsequently shed. Speyer (23) also found tetrachloroethane effective against adult white flies, but its effect on the red spider was only temporary. The tetrachloroethane, when used at the rate of 10 ounces per 1,000 cubic feet, severely injured several varieties of chrysanthemum. Richardson and Smith (21) found that 34 and 31 grams per 100 cubic centimeters of chloroform and carbon tetrachloride, res- pectively, were necessary to kill aphids, and that the tolerance of the nasturtium plant for these two compounds varied from 5 to 8. Chlorobenzene and commercial trichlorobenzene were less toxic to aphids than carbon disulphide, and their toxic concentrations ex- ceeded the tolerance of the plant. Moore (14, 15) used chlorobenzene, p-dichlorobenzene, chloroform, carbon tetrachloride, and chloretone (trichloro tertiary butyl alcohol) in various tests on flies. FLUORIDES Fluorobenzene at a concentration of 1.9 per cent showed 100 per cent mortality against Sitophilus oryza, but diflu or o diphenyl ex- hibited almost no toxicity, owing probably to its slight volatility at the temperature of the test. Organic fluorine compounds have not been tested by many investi- gators, probably because they are ra.re and expensive. IODIDES The only iodide tested, normal butyl iodide, killed all the weevils at a concentration of 0.8 per cent. Moore (14) found iodobenzene to be more toxic to house flies than the corresponding bromine and chlorine compounds. Tattersfield and Roberts (24) found iodoform to be nontoxic to wireworms, while iodobenzene was moderately toxic. According to Holt (9), cockroaches dusted with iodoform did not succumb until a period of 9 hours had elapsed. ALCOHOLS AND PHENOLS Methyl alcohol was more toxic than ethyl alcohol, but with this exception the toxicity increased with increasing molecular weight through isoamyl alcohol. Thymol, menthol, and geraniol are so slightly volatile at ordinary temperature as to be practically nontoxic. Richardson and Smith (21) found methyl, ethyl, n-propjl, n-butyl, capryl, isoamyl, benzyl, and furfuryl alcohols to be ineffective against aphids. Even pure methyl and ethyl alcohols killed less than 95 per cent of the insects. Moore (15) found the toxicity of methyl, ethyl, and amyl alcohols, menthol, and thymol on house flies to increase with decreasing volatility. Trillat and Legendre (26) showed that methyl, ethyl, propyl, and amyl alcohols had only a feeble toxicity to mosquitoes. 4 Sycamore maple. FUMIGATION AGAINST GRAIN WEEVILS 15 Titschack {25) showed that ethyl alcohol was ineffective against the eggs and larvae of clothes moths. McClintock, Hamilton, and Lowe {IS) found the effectiveness of methyl and ethyl alcohols to he low against bedbugs, cockroaches, house flies, clothes moths, and mosquitoes. Menthol was about half as toxic to bedbugs as carbon disulphide. Holt {9) found methyl alcohol to be more toxic to cockroaches than ethyl or amyl alcohol. All required at least 45 minutes to produce death, however. Lefroy {11) reported that 70 per cent ethyl alcohol failed to kill any mealworms dipped in it. Burmeister in 1836 {3, p. 39) recorded instances of beetles which after being immersed in spirits of wine for 12 hours recovered all their functions when removed from it. Apparently alcohols of the fatty acid series have low toxicity to insects and are not suited for fumigants. ALDEHYDES Crotonaldehyde was much more toxic than n-b u ty r al d ehy d e . Furfural and benzaldehyde showed marked toxicities at concentra- tions of less than 1 per cent. The chlorine-substituted aldehyde, chloral hydrate, had low toxicity. Richardson and Smith {21 ) tested paraldehyde, aldehyde ammonia, chloral hydrate, furfural, and benzaldehyde against aphids. Even benzaldehyde, the most effective, was not a practical aphicide, as a 5 per cent solution was required to kill. None of the many experimenters with formaldehyde, an effective fungicide, has discovered any practical value for it as a fumigant against insects. Phelps and Stevenson {20) found a 0.5 per cent solu- tion to be effective as a stomach poison to flies, possessing a co- efficient of 2.32 as compared with a coefficient of 1 for 0.001 normal arsenite solution. Four and 8 per cent formaldehyde solutions were less effective than the 0.5 per cent solution; a 1 per cent solution had a coefficient of 2.36. The addition of molasses and brown sugar diminished the effectiveness of the formaldehyde solutions. Dry powdered paraformaldehyde showed a coefficient of only 0.14, while a saturated solution had a coefficient of 1.90. Moore (15) found that the toxicity of acetaldehyde, chloral hydrate, and furfural to house flies in general increased as their volatilities decreased. According to Davis {5), acetaldehyde is ineffective against white grubs in soil. Trillat and Legendre {26) report that it has only a feeble toxicity for mosquitoes. Lefroy {11) found a solution of chloral hydrate in water to have no effect on mealworms. Holt {9) reports that roaches dusted with powdered chloral hydrate survived for 4 hours. McClintock, Hamilton, and Lowe {IS) found benzaldehyde to be more effective than carbon disulphide against bedbugs, cockroaches, and house flies, equally effective against clothes moths, and less than half as toxic to mosquitoes. • KETONES Ethyl methyl ketone was more effective than acetone (dimethyl ketone). The introduction of chlorine into a ketone, as in cliloro- acetone, enormously increased its toxicity. 16 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Richardson and Smith (21) found that even 100 per cent acetone and ethyl methyl ketone killed less than 95 per cent of aphids, and acetal was effective only in concentrations over 50 per cent. Moore (15) found the toxicity of acetone, bromomethyl phenyl ketone, and menthone to house flies to increase with diminishing volatility. Holt (9) found that it took acetone vapor 32 minutes to kill cock- roaches, as against 8 minutes for carbon disulphide. Lefroy (11) found that acetone killed no mealworms dipped in it. According to McClintock, Hamilton, and Lowe (13), acetone has very little effect against bedbugs, cockroaches, house flies, clothes moths, and mosquitoes. Titschack (25) found acetone less effective than carbon disulphide against the eggs and larvge of clothes moths. Guerin and Lormand (7) found that bromoacetone had no apparent effect on plants at a concentration of 1 to 2,000 exposed for an hour. Bertrand and Rosenblatt (2) found monochloroacetone to be very much more active against the larvae of (Bombyx) Malacosoma neustria L. than ether, chloroform, carbon tetrachloride, or carbon disulphide, but less active than hydrocyanic acid or chloropicrin. ESTERS The esters showed a higher toxicity to weevils than any other class of organic compounds tested. Methyl, ethyl, isopropyl, and methyl cyano formates and isoamyl acetate are more toxic, molecule for mole- cule, than carbon disulphide. Propyl acetate is less toxic than car- bon disulphide; ?i-butyl acetate, isobutyl acetate, and ethyl-n- valerate are so slightly volatile at the temperatures used that they do not give 100 per cent mortality in the concentrations obtained. Richardson and Smith (21) tested the action on aphids of amyl acetate, methyl salicylate, and benzyl acetate. Benzyl acetate, at a concentration of about 1 per cent, was the most effective. Moore (15) tried the effect on house flies of methyl salicylate, ethyl malonate, ethyl acetoacetate, amyl acetate, amyl valerate, and propyl acetate. Lefroy (11) found that amyl acetate killed all mealworms dipped in it, while methyl salicylate and ethyl acetate killed some. Moore (16) tried the effect of a wide variety of substances on the clothes louse. He concluded that liquids less volatile or more volatile than creosote are not as successful, and that, while not quite as good, crude phenol and methyl salicylate are the best substitutes. ETHERS s-Dichloromethyl ether had an efficiency comparable to that of carbon disulphide; chloromethyl ether was about half as effective. Acetal was effective at a concentration of slightly over 4 per cent. Ethyl ether was the most volatile and least toxic of all materials used by Moore (15) against house flies. Ethyl ether has been tested by a number of other experimenters, but in all cases it showed only a low toxicity against insects. CHLOROHYDRINS Epichlorohydrin, the most toxic substance tested in this investiga- tion, killed 100 per cent of S. oryza at a concentration of only 0.09 per cent. Ethylene chlorohydrin at the maximum concentration tested, 1.4 per cent, was ineffective. FUMIGATION AGAINST’ GKAIN WEEVILS 17 Apparently chlorohydrins have not been tested by other experi- menters with insecticides. SULPHUR COMPOUNDS Methyl and ethyl thiocyanates and allyl isothiocyanate were more effective than carbon disulphide; methyl and ethyl sulphides and ethyl mercaptan were less effective than carbon disulphide. Butyl mercaptan was as effective as carbon disulphide against S. oryza, but only half as effective against the other species of weevils. Mercaptol is so slightly volatile as to be ineffective. Cyanogen sulphide killed no weevils at a concentration of 0.1 per cent; per- chloromethylmercaptol and p-toluenesulphochloridegave variable kills. Richardson and Smith (21) found that a concentration exceeding 5 per cent of carbon disulphide was necessary to kill aphids, whereas a solution containing less than 2 per cent injured the nasturtium plant. According to Moore (15), molecule for molecule, allyl isothiocyanate is more toxic against house flies than chloropicrin, and carbon disul- phide and ethyl mercaptan are more toxic than their relative vola- tilities would indicate. Tattersfield and Roberts (24) found allyl isothiocyanate to be the most toxic to wireworms of all compounds tested by them. The toxicity of carbon disulphide was equal to that of benzene. Speyer (23) found ethyl mercaptan to have no effect on red spiders. It also failed to kill adult white flies. NITRILES AND ISONITRILES n-Valeronitrile had a toxicity as great as that of carbon disulphide;, acetonitrile was about one-fifth as toxic as carbon disulphide. Pro- pionitrile and phenylisonitrile were ineffective. Phenylacetonitrile was the most toxic of this class of compounds, killing all S. oryza at a concentration of 0.10 per cent. Moore (14) found that the toxicity of benzonitrile to house flies com- pared with that of iodobenzene and xylene. NITRITES n-Butvl nitrite and isoamyl nitrite were about equally toxic. Moore (15) found that amyl nitrite had about the same toxicity to flies as gasoline. Tattersfield and Roberts (24) found that the toxicity of amyl nitrate to wireworms was low and that the toxicity of amyl nitrite was moderate. Speyer (23) found that the grubs of a chironomid fly (Orthocladius) came to the surface of the soil of cucumber pot plants and were killed in a short time when amyl nitrite and amyl nitrate were used. He also found that methyl nitrite had a more permanent effect on red spiders than amyl nitrate and tetrachlorethane, but that it was necessary to use concentrations which hurt the plants. The adults, of the white fly were killed by all these vapors. NITRO COMPOUNDS Nitromethane was the only nitro compound which killed all of the insects ; a concentration of 3.6 per cent was necessary. Nitrobenzene, p-chloronitrobenzene, m-dinitrobenzene, and a-nitronaphthalcnc at saturation concentrations killed none of the weevils. Moore (15) found nitrobenzene even more toxic than nicotine to house flies. 15085° — 25f 3 18 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Several nitro compounds have been, tested against wireworms by Tattersfield and Roberts (24). m-Nitroaniline, o-nitro aniline, p-nitro- phenol, nitronaphthalene, dinitrobenzene, and nitrobenz aldehyde were nontoxic; o- and p-nitrochlorobenzene, o- and p-nitro toluene, p-nitro aniline, and nitroxylene (mixed derivatives) had marginal toxicity; nitromethane had. low toxicity; nitrobenzene had moderate toxicity; and o-nitrophenol and nitrochloroform (chloropicrin) had high toxicity. AMINES Diethylamine and n -butyl amine were more toxic to weevils than carbon disulphide. None of the other amines tested (aniline, methylaniline, dimethyl aniline, ethylbenzylaniline, acetphenylene- diamine, and a-naphthylamine), however, showed any toxicity. Richardson and Smith (21) tested the action of the following com- pounds on aphids: Trimethylamine hydrochloride, tetramethylam- monium chloride, p-phenylenediainine, phthalimidine, methylamine hydrochloride, diethylamine, triethylamine, tri ethyl amine hydro- chloride, tetraethylammonium chloride, tetrapropylammonium hydroxide, isobutylamine, diamylamine, triacetonamine, hexamethyl- enetetramine, formamide, dicyanodiamide, choline hydrochloride, betaine hydrochloride, nitroguanidine, succinimide, aniline, benzyl- amine, benzidine hydrochloride, m-phenylenediamine hydrochloride, camphylamine, and tetrahydrobetanaphthylamine hydrochloride. None of these approached nicotine in toxicity. Tetramethylam- monium chloride, the most effective, required a concentration of 0.35 per cent, compared with an effective concentration for nicotine of only 0.007 per cent. Tattersfield and Roberts (24), in tests on wireworms, found that o- and m -nitro aniline, w-phcnylenediamine, phenylhydrazine, naph- thylamine, and diphenylamine were nontoxic; that p-nitroaniline and p-chloro aniline had marginal toxicity; that trimethylamine, ethvlamine, dimethylamine, monomethylamine, aniline, and o-chloro- aniline had moderate toxicity; and that o- and p-toluidine, xylidine, dimethylaniline, and monomethylaniline had high toxicity. Foreman and Graham-Smith (6, p. 113) found aniline (saturated aqueous solution) and aniline hydrochloride to be toxic to flies when taken by the mouth. The hydrochloride of o-toluidine had a similar effect, but p-toluidine hydrochloride was nontoxic. Flies taking monomethylaniline hydrochloride appeared to be dead in 10 minutes, but recovered in 2 hours. Hydroxylamine hydrochloride in 2 per cent solution had little effect, or none, and a 1 per cent solution of m-phe- nylenediamine had no effect. Jewson and Tattersfield (10) tested the action on mites ( Aleurobius farinse) of aniline, monomethylaniline, and dimethylaniline. Most of the large mites remained alive, but moved sluggishly. MISCELLANEOUS NITROGENOUS COMPOUNDS Pyridine at a concentration of 0.8 per cent killed all the weevils, but hexamethylenetetramine showed no killing power. Richardson and Smith (21) sprayed aphids with the following pyridine and quinoline derivatives : Nicotinic acid nitrate, 4-dimethyl- aminoantipyrine, n-ethyl piperidine sulphate, n-ethyl piperidine, crude chloropiperidine, piperidine sulphate, pyridinium ethyl hydrox- ide, pyridine, pyridinium ethyl iodide, piperidine, methylene dipi- peridine, 7 7 dipyridyl, a-picoline, pyrrole, quinoline, tetrahydro- FUMIGATION AGAINST GRAIN WEEVILS 19 quinoline, quinaldine, and piperazine. The minimum toxic concen- tration of pyridine was 25 grams per 100 cubic centimeters of solu- tion. None of these compounds was more than one-hundredth as effective as nicotine. McClintock, Hamilton, and Lowe (13) found little difference in the toxicity of pyridine, pyridine bases, and quinoline for bedbugs, cockroaches, house flies, clothes moths, and mosquitoes. Jewson and Tattersfield (10) found pyridine effective against mites. Lefroy (11) found that quinoline and pyridine killed some of the mealworms dipped in the liquids, but that a 1 per cent aqueous solution of nicotine -killed none. Tattersfield and Roberts (24) found that pyridine was moderately toxic against wireworms. Moore (15) found pyridine somewhat more toxic than its position in the table of organic compounds arranged according to volatility would indicate. Its toxicity was about the same as that of furfural. INORGANIC COMPOUNDS None of the inorganic compounds tested proved suitable as a fumigant. Thionyl chloride killed all weevils at a concentration of 4.5 per cent; concentrations of carbon dioxide ranging from 50 to 100 per cent were required to effect a complete kill. RELATION BETWEEN VOLATILITY AND TOXICITY OF FUMIGANTS The compounds causing 100 per cent mortality to S. oryza (selected on account of its high resistance to fumigants) exposed to them for 24 hours, together with their boiling points, are listed in Table 3. Table 3 .— Toxicity of fumigants to Sitophilus oryza ( arranged in order of decreasing effect ) Fumigant Minimum concentra- tion caus- ing 100 per cent mor- tality in 24 hours Boiling point 1 Minimum quantity causing 100 per cent mortality in 24 hours Cost per pound 2 Cost per 1,000 cubic feet Per cent i 0. 09 °C. 116. 6 Lbs. per 1 ,000 cu.ft. 0. 23 Dollars 27. 22 Dollars 6. 26 . 10 231. 7 .33 18. 14 5. 99 . 10 247. 0 .43 9. 07 3. 90 . 10 133. 0 . 20 27. 22 5. 44 3 . 16 202. 0 . 54 5. 90 3. 18 .26 119. 0 .67 27. 22 18. 24 3 . 27 178. 3 .80 1. 13 .90 3.34 150. 7 .94 4. 08 3.84 3 . 50 104. 0 .98 .50 129. 0 2. 62 .91 2. 38 3 . 50 70. 0 1. 23 2. 27 2. 79 . 55 130. 0 1. 50 13. 61 20.42 3.60 99. 0 1. 96 9. 07 17. 78 .60 54. 4 1. 24 .40 . 50 .60 75. 0 1. 72 3. 63 6. 24 .60 76. 0 1. 22 36.29 44. 27 .68 100. 0 1. 61 3.70 96. 8 2. 20 1. 27 2. 79 3.80 161. 0 2. 14 . 25 . 54 .80 74. 1 2. 97 2. 27 6. 74 .80 131. 0 1.96 1.81 3. 55 n-Butyl iodide .80 129. 9 4. 10 9. 07 37. 19 Ethyl thiocyanate .80 146. 0 1.94 22. 68 44.00 Pyridine .83 116.0 1.83 .36 .66 1 Boiling points taken from Beilstein. iMost of the prices are taken from List No. 10 of the Eastman Kodak Co., Rochester, N. Y., issued September, 1923; some are from wholesale quotations by the Oil, Paint, and Drug Reporter, March 3, 1924; arid some are quotations from the Miner Laboratories, Chicago, 111. 2 Minimum concentrations tested. 20 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Table 3. — Toxicity of fumigants to Sitophilus oryza ( arranged in order of decreasing effect ) — Continued Fumigant Minimum concentra- tion caus- ing 100 per cent mor- tality in 24 hours Boiling point Minimum quantity causing 100 per cent mortality in 24 hours Cost per pound Cost per 1,000 cubic feet Methyl formate Per cent °C. Lbs. per 1,000 cu.ft. Dollars I Dollars 3.90 32.3 1. 51 1. 59 2.40 Isoamyl acetate .90 139.0 '3.26 .91 2. 97 Diethylamine .90 55. 0 1. 83 16. 33 29.88 m-Butyl mercaptan ... . .90 97.0 2. 26 36.29 82.02 Bromoform ..... 3.94 150. 5 6. 62 2. 04 13. 50 o- and p-Dichlorobenzene 1. 00 175. 0 4. 10 .15 .615 Carbon disulphide 1. 00 46.0 2. 12 .06 . 127 s-Tetrachloroethane 1. 10 147.0 5.15 1. 13 5. 82 . zi-Valeronitrile . 1. 10 141.0 2. 55 27.22 69.41 s-Dichloromethvl ether. 1. 13 105.0 3. 62 27. 22 98.54 w-Propyl alcohol 1. 70 97.4 2.85 1. 36 3.88 m-Butyl alcohol .. 3 1. 70 117.0 3. 51 .76 2. 67 Ethyl methylketone .. Fluorobenzene.. 3 1. 80 3 1. 90 79. 6 84. 9 3. 62 5. 09 .30 1.09 Chlorobenzene 2. 00 132.0 6. 27 .09 . 58 Chloromethyl ether ..... 3 2. 20 59. 5 4. 94 13. 61 67. 23 ■71-Propyl acetate.. ... 2. 50 101. 6 7. 11 2. 27 16. 14 -71-Butyl bromide 2. 60 101.0 9. 93 4. 54 45.08 Tetrachloroethylene 2. 60 121. 0 12. 02 1. 13 13.58 Ethyl sulphide 2. 90 92. 0 7.29 2. 27 16. -54 Ethyl bromide. . . . ... 2. 90 38. 4 8. 81 .40 3. 52 Nitromethane ... 3. 60 101. 0 6. 12 9.07 55. 51 Benzene 3.80 80. 2 8.27 * .04 .33 Allybbromide ... .. . 3 3. 90 70. 0 13. 15 15.88 208. 82 -71-Propyl bromide 4. 10 70.8 14. 05 13. 61 191.22 Acetal 4.30 102. 2 14. 15 18. 14 256. 68 Methylene chloride... . 4. 40 47. 6 10. 42 6. 80 70.86 Thionyl chloride _ 3 4. 50 78.0 14. 92 9. 07 135. 32 Acetonitrile .. ... ._ .. . . . . . .. 5. 00 81. 6 5. 71 18. 14 103. 58 ■n-Butyraldehyde 5. 50 77.0 11. 05 13.61 150. 39 Ethylene chloride 6. 00 83. 5 16. 55 .68 11.25 Carbon tetrachloride 6. 90 76.7 29. 58 .082 2.44 Chloroform . .. 7. 00 61. 2 23. 29 .32 7. 45 Acetone. .. 8.20 9. 00 56. 5 55. 0 13. 27 24. 31 .20 2.65 Trichloroethylene 10. 00 88.0 36. 62 .36 13.18 Ethylidene chloride 10. 20 59. 2 28. 13 22.68 637.99 Methyl sulphide _ 10. 40 38.0 18.00 45.36 816. 48 Cyclohexane .. .. 3 13. 00 80.8 30.48 22. 68 691.29 Methyl alcohol 16. 00 64.7 14.28 « . 15 2. 14 Ethyl mercaptan.. 3 19. 00 37. 0 32. 89 9. 07 298. 31 Amylene 40.30 22-37 78. 76 9. 07 714. 35 Carbon dioxide 50. 00 -78.5 61. 32 .07 4.29 s Minimum concentrations tested. 4 Chemically pure, in drums, sold for 30 cents a gallon. s Purified, in drums, sold for $1 a gallon. While the volatility of organic compounds at ordinary temperatures is not proportional to their boiling points, very few data on the vapor pressure of these compounds at ordinary temperatures are available. For that reason the boiling points were used in this investigation. The compounds are arranged in the order of decreasing toxicity. Moore (15) states: In general, the toxicity of a volatile organic compound is correlated closely with its volatility. A decreasing volatility is accompanied by an increased toxicity. The boiling point of the chemical is a general index of its volatility. Compounds with boiling points of 225° to 250° C. are usually so slightly volatile that they do not produce death except after very long exposures. If this theory is true, the boiling points of the compounds should show a decrease, indicating increasing volatility, but no well-defined relation between the toxicity and the boiling point is shown in Table 3. This is not in agreement with Moore’s theory. Table 4. — Results of fumigation tests on weevils in grain in 19-liter glass bottles FUMIGATION AGAINST GRAIN WEEVILS 21 Weevils killed after exposure for 24 hours Top of bottle Tribo- lium Per cent 100 100 100 90 100 100 100 20 :ggggggggggggggg : IHHHHHHHHHHH h h h ' 1 1 jggggggggg 05 2.g OQ ° Per cent 100 50 25 0 100 0 50 25 ! N 1 1 J i : j i ; ; i i js j i i i i i i i i i i i i i i i i i i S. oryza Per cent 100 90 100 75 100 100 100 iggcsggggSgKoKgc i j — .100 100 0 100 100 100 100 100 100 Middle of bottle Tribo- lium Per cent 100 i i i i i i ls222l°lll§0§sl : : jggggggggg j S. gran - arius Per cent 100 : ; is j S. oryza Per cent 100 |i j | ] j- ggKgKg°gg£g°gg j igg°gggg|o Bottom of bottle Tribo- lium Per cent 100 00 100 100 100 100 60 100 2S2222222222°^2® llgggggggg S. gran- arius Per cent 100 0 0 25 100 25 75 0 100 .(III! 1 i i i i i i , . i !S : j ; 1 1 • • 1 • 1 • 1 S. oryza |o°g«ooS0§gooNoggoN§o§°gc : : Dgg°gggggg Average mortal- ity l§888§§i23§§§§£§ggS§§ggSfegggS§8388§88§ a, Fumigant §§ g'-S |JS Lbs. per 1,000 cu. ft. 10.0 10. 0 15.9 8. 0 10. 0 10. 0 15. 9 8.0 11. 25 15. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 10. 0 ; j cc ; OOOOOOOOOO 1 doco'ddddddo j ; i Material Carbon tetrachloride ao_ do do do do uu do do do do do do do do N §■§ : : 1 Carbon tetrachloride j j j j j j j j j goooc'odo i j j j|| j j j j If Lbs. per 1,000 cu. ft. 0. 35 .35 .47 .67 1.00 1.00 1. 25 1.50 3.75 5.00 3. 10 5. 20 3. 10 6. 20 4. 70 2.90 2. 90 3. 40 4. 90 4.90 2. 90 5. 58 11. 10 1 70 .26 .52 1.03 2. 06 2. 06 2. 50 3. 10 9. 10 18. 20 Material Methyl thiocyanate _. do au Chloroacetone do do_ do Turpentine.. An Ethylene bromide.. 72 -Butyl nitrite do do Isoamyl nitrite.. Crotonaldehvde i i i i : i |! i jo Ml h ji : i JI i ; : :■ i i i i j i j Epichlorohydrin do do do Ethyl formate... do Ethyl bromide i ! i ! ; i i i® IsgiIlslsls8Sli'$3SSJS8SfigaiI?llsSss Table 5. — Results of fumigation tests on weevils in boxes 'partially filled with grain 22 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE E!P°Id mmpqijx snuvuviO -g mfUo -g oooooo *-* 000000 >000000 tnpou nmnoquji -oooooo ^oooooo ooooooo o o o o o snuvuvifi ’g O o o o o o o £> O O O O 0Z/SiO ‘S' >000000 J o o o o o o > ^2^ c5cOCOo6 cOtO^COCOOOO'O^^HH CO CO* CO CO* CO* 73 j t3 Ti 5-1 ’7"? **~ 1 . ® S-i S oS 0 73.2 073 O 9 3 9 3 '3 ■93 ‘3 3 3 23 g 23 £ £ g 3-23-2 1,0 T) ■ ■ ' a-0 a a-a a ' ^ ! ! i r-i O O ! >573 73 5 :s >0^0 3X35'° ! ! >73 jrd J :S ,r C3* O ^>73 n co o NCOOHCIO lOCO CO «5 lO LO Cl CO 1-H r-i !W NCOCONOONCOOJ'fvOOSiOH^NOOOH l,“ ^ ' 'NNOCCOCOCCCSOiQOOOOONN iO IQ lO lO O CO O CO co CO i0i0C0c0c0L010O‘0i0i0‘0i0*0‘0‘0i0i0 FUMIGATION" AGAINST GRAIN WEEVILS 23 o o o o oo OOOCOIMOIOOO 05000100 8 8 § § .0*00000 o o o o o o o o oooooo i O O O CO *o o o oooooo o oooooo .000000*000 o 10000 or- oo I O O O O O O I o o o o o .0000*00 O COOO--COOONCOOiO!NtJooooor-C'**o*OH»o o r- co o o*ooocd o co*oo*or-co O GO O O *0 O CONOHOJN 'CO— >— ‘COCO oi cd cd cd ci oi oi co co OCOOCOOCOOCOOCO 0*00*00*00*00*0 o.S o 3 as O O © JD I C3 C3 i ® • © • © i J X3 X3 [ 'O J cd "c © -n © ‘c © X3 0X3 0X3 0X3 PI^PI O G3 O C3 O co o - ©-S ©-S ©-S © d /-i d «-* d pH d o 9 © S © 9 © - 0—1 O — ' o^ ^^COCDOiO co cd *o o c5 tji :3 . ; ©-a is a i © o I 4-3 *-* . d^ o o O O 0—1 ^X3x3x3jO >» ■ ■ ■ !8S low ‘OCOCOOlOlOOO-^tCOO' cdcdcd— i-^ooocricit--}i' *o CO 5 05 GO GO O * oi *o‘ *d o CO CO o 0*0 10 000 •O d^5 K C3 M w S'CSk flja a o •g o a- o a» « « o n • j-l J.§ * a o o £ d o SaJocit-i I'S. 0 a O O IW l-C-C u :S go C5 oooc-"Mp5P?oonu5noh^c« co -r lo It It » ' i - to -r Oi o> co oi — « co co o5 go O go c. o> 05 cb c& O o co h *0 *0 »0‘0*0*0*0'.0'.0'.0‘0'?5'--— ' *o *0*0 *0 CO CO cO CO 1 Same results obtained for tests Nos. 604, 631, and 632. 24 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE EFFECT OF FUMIGATION ON WEEVILS IN THE PRESENCE OF GRAIN IN BOTTLES, BOXES, AND BARRELS In order to ascertain the action of the compounds as they would be used in practical fumigation work, weevils were placed in pill boxes, with perforations to permit the entrance of the vapor, and the boxes were placed at different levels in grain, usually wheat. The weevils, usually about 10 to the box, were exposed to the action of the com- pounds for 24 hours, at temperatures from 21° to 32° C. One series of tests was carried out in large glass bottles of 19 liters capacity (Table 4), another in a wooden box of 100 cubic feet capacity (Table 5), and another in a barrel of 6 cubic feet capacity (Table 6) . The fumigant or mixture of fumigants was applied in all these tests by pouring the calculated quantity over the top of the grain and then closing the receptacle. The bottle had a stopper, the barrel had a wooden head, and the box was provided with felt strip- ping and a tight-fitting cover, which could be clamped down. (In Tables 4 and 5 the more toxic substance, with its concentra- tion, is given first; the less toxic substance, used principally as a diluent, with its concentration, is given next.) Table 6. — Results of fumigation tests on weevils in grain in barrels ( 6 cubic feet capacity ) 1 Barrel made as tight as possible in this test. The insecticidal action of a gas is greatly lessened by the presence of grain, probably because the grain absorbs many vapors in large quantities and because the grain mechanically interferes with the diffusion of the gas throughout the receptacle. For example, in a glass vessel containing nothing but weevils and the vapor of the compound (mixed with air), epichlorohydrin killed 100 per cent of the insects at a concentration equivalent to 0.23 pound per 1,000 cubic feet; when the weevils were planted in wheat, a concentration FUMIGATION AGAINST GRAIN WEEVILS 25 of epichlorohydrin equivalent to 2.06 pounds per 1,000 cubic feet, combined with carbon tetrachloride at the rate of 10 pounds per 1,000 cubic feet, did not kill all of the insects in every test. The compounds consistently causing 100 per cent mortality in the small-scale tests in wheat (Tables 4, 5, and 6) are shown in Table 7. Table 7. — Most effective fumigants f or weevils ( based on results in Tables'4, 5 , and 6) Test No. 456 462 629 630 615 616 611 612 599 600 540 545 585 586 Fumigant j-Ethyl bromide . 1 Ethyl bromide / Carbon tetrachloride. { Ethyl bromide Carbon disulphide _ . - Carbon tetrachloride. (Ethyl bromide Carbon disulphide . Carbon tetrachloride. Ethylene bromide ... 1 Ethyl bromide / Ethylene bromide. . . 1 Ethyl bromide /Methyl formate 1 Ethyl bromide /Methyl formate Mini- mum concen- tration consist- ently causing 100 per cent mor- tality of all species after ex- posure for 24 hours Lis. per 1,000 cu.ft. 18.2 4.5 10.0 4.5 2.2 10.0 4.5 2.2 10.0 5.3 13.2 5.3 1.7 6. 1 11.2 3.5 Test No. 439 603 604 631 632 589 590 594 597 598 472 546 547 Fumigant f Ethylene bromide . . . \Carbon tetrachloride. I Ethylene bromide . . . (Carbon tetrachloride . \ Ethylene bromide _ . . / Carbon tetrachloride. [Ethylene bromide Chlorobenzene Carbon tetrachloride . /Crotonaldehyde (.Carbon tetrachloride. / Ethyl formate ICarbon tetrachloride. | Carbon disulphide. Mini- mum concen- tration consist- ently- causing 100 per cent mor- tality of all species after ex- posure for 24 hours Lis. per 1,000 cu.ft. 6.2 10.0 8.0 8.0 5.3 12.0 5.3 10.0 10.0 2.9 10.0 2.5 10.0 6.4 With the exception of carbon disulphide, the most economical fumigant at 1924 prices, is the combination of ethyl formate with carbon tetrachloride. (The ethyl acetate-carbon tetrachloride mix- ture,5 which was the most suitable and economical fumigant of all those tested, was not tried on a small scale.) IN BOX CARS Many tests were also carried out in box cars of m-ain. Weevils in perforated pill boxes (10 in each box) were planted at 10 different levels in the grain in a box car, by putting the boxes in the compart- ments of a grain sampler, which was then plunged into the gram at an angle of about 45°. The fumigant was applied to the grain by sprinkling the liquid over the surface from a small sprinkler-top watering can as quickly as possible. The door of the car was then closed and sealed. Twenty-four hours later the pill boxes were withdrawn and the number of dead weevils was determined (Table 8). (The more toxic substance, with its concentration, is given first; the less toxic sub- stance, used principally as a diluent, with its concentration, is given next in Table 8.) * The use of ethyl acetate and ethyl acetate-carbon tetrachloride mixture was suggested by K. O. Roark. Table 8. — Results of fumigation tests on weevils in wheat in hox cars 26 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE FUMIGATION AGAINST GRAIN WEEVILS 27 28 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE Many of the fumigants and combinations of fumigants which operated successfully in glass jars and in the presence of wheat in bottles, boxes, and barrels gave disappointing results when used on a large scale. A mixture, in equal proportions by weight, of ethylene bromide and carbon tetrachloride, used at the rate of 16 pounds per 1,000 cubic feet, was successful when applied to weevils in wheat in a box, but this mixture, even with the addition of 50 per cent more carbon tetrachloride, was ineffective when used in box cars. This is probably explained by the facts that in a box car the weevils may be in the grain at a much greater depth than in a small box or barrel and that the grain is much more tightly packed. Also, a box car is not a tight container and the vapors of the fumigant may not be long retained. The most successful fumigants used in the box-car tests were ethyl formate in combination with carbon tetrachloride (12.1 pounds each per 1,000 cubic feet) and ethyl acetate (12.5s pounds per 1,000 cubic feet) in combination with carbon tetrachloride- (25 pounds per 1,000 cubic feet). FIRE HAZARD FROM FUMIGANTS The vapor pressures of ethyl acetate and carbon tetrachloride are very close for all temperatures up to their boiling points (Table 9, Fig. 1). Table 9.— Vapor pressures of ethyl acetate and carbon tetrachloride Temper- ature Observed vapor pressure 1 Temper- ature Observed vapor pressure 1 Ethyl acetate 2 Carbon tetra- chloride3 Ethyl acetate 2 Carbon tetra- chloride3 °C. mm. mm. °C. mm. mm. -20 6. 55 9. 92 40 186. 20 210. 90 -10 12. 95 18. 81 50 282.20 309.00 0 24. 30 33. 08 60 415. 40 439.00 10 42. 70 55. 65 70 596. 30 613. 80 20 72.80 89. 55 80 832. 70 < 836. 35 30 118. 70 139. 60 1 Sydney Young. The vapor-pressures, specific volumes, heats of vaporization, and critical constants of 30 pure substances. In Sci. Proc. Roy. Dublin Soc., n. s. (1909-10), 12:374-443. 2 Boiling point at 760 millimeters, 77.15° C. 3 Boiling point at 760 millimeters, 76.75° C. 1 Calculated. A mixture of 3 volumes of carbon tetrachloride and 2 volumes of ethyl acetate (equivalent to 72.5 per cent carbon tetrachloride and 27.5 per cent ethyl acetate by weight) is noninflammable at ordinary temperatures. Moreover, in this mixture the vapors of the two components tend to separate but very slightly, thus making the mixture safe from fire hazard. On the other hand, carbon disulphide has a much higher vapor pressure than carbon tetrachloride at ordinary temperatures, thus making mixtures of these compounds unsafe. The experiments of the Underwriters Laboratories at Chicago (Grain Dealers Journal, December 10, 1921, vol. 47, p. 798) show that a mixture of 75 per cent carbon tetrachloride and 25 per cent carbon disulphide by vol- 6 A mixture of ethyl acetate and carbon tetrachloride containing 33§ per cent by weight of the acetate will flash slightly at ordinary temperatures but will not continue to burn. FUMIGATION AGAINST GRAIN WEEVILS 29 mne (equivalent to 79.1 per cent and 20.9 per cent by weight) gives off a highly inflammable vapor. The use of such a mixture to kill weevils or other insects is condemned as dangerous. A mixture of carbon tetrachloride and carbon disulphide containing as little as 5 per cent by volume of carbon disulphide will flash at 20° C., although it will not continue to burn. Fig. 1.— Vapor pressures of ethyl acetate and carbon tetrachloride (Tabic 0). 10.— Results of milling tests with wheat exposed to fumigants 30 BULLETIN 1313, U. S. DEPARTMENT OE AGRICULTURE a * 03 >, £ a t*£ P “Isa •s s fe 3 P t> ho t*>>g ° P, .g fe 03 a * SS <03 •*- fl 42 ca O s-« . w to J'd'dJ" s a SS o . ox) *5 3 42 o o *-. rO J_l I ©' d * a 53 s o o a -*a aJ co ai o |g a I §gla- bfl°< g ■“ s a § O tj3 HI 13 03 O © 03 P P ,© g sg 3 3 s s s — O co .. _ ©,a 2^ t>»2 >»2 g.sfo :fe.£fs§ &3Z l>53!>!? ^ W O tuG^ _ J3 >>’0 SjsO •• 1 S§ ! s mS 'i% ® J?a2> ao” °iio £ CQ> •S? 6 Eo£ 33 .BPS Sfc 1 r 1 ojj 1 1 qu > > 1 +3 ii: 1 1 ]zi ' d O co bfico MO CO J S o t»g >,§13 >,.£3 3 1 t. P >_ g I_ bfi-*3 >m |>gq55 -a 02Z !>S ^5 <5 >£ bo >> >> M "3 w w OJU JJ QU DU cu OU P. DU 1»3 >>1 gSgSgo g I g -£* g fe .S1 fe ® 2 ^2“£-a 2 s 2 oS 2 5 > gq > ^ co Thuimm'A m %, cq 2 go O OOOOU0U0OO COCOCCO>COCX)OiO Tji c Enpe :S ■gis o3 OP is a .Sis .Sis ;« ;« :«is uT3£®'dg®i3-Si3®’a®® Sgoj^o'S— fl'Srt'So3’c3o3g !>>,o £ >>j3 2”>>j3 £1*i S’?! SaaSSSSaSSSSaStf w ' 03 a>'d'£ <^T3 cj +3 a? o ■+■* © 03 gS 03 g« >-< O 07^ O <13 P^ - .pw - ^ . r- r . - >>pQ SaaSc3o5rto5a5o5c5o3feci OP H DO HOP pd^)p£||D BPS H OjD ^t^oooiO'^oocqeoi-Huo oooojojoooooooco o • -g.S os3 S§ a h o o o o o 0 -C 'O XI "O >>3 .q -c-a >>3jo 'O ■ 'So& I > ogoooo>oo»oo . >c d >o o d io oi io C“1 LO r-< ^0 M CM -H o o o o o o o o o ico icoio ooo o »o o s °-a >"9 &,a US S't’O'b g'S-Ob §3 O. II w.fl ~ —> Cj SZ is ■ I/S'-'S S © fl,2 c.2 w.fl —'.fl w,p 2*3%3%2%2% ^ ^ ^ h» ^ 2 ^ 3 |s|s g2f2|2'§'8”r O^flO-flOlp'Sfl^fl'^CJp' — cr c- O'- C-H G G-r O >.3 -'..G >*jD >.,G >..G i- aSSSSoSa' WOPOMOWOWOPHO! p.03 t- dJ o © t. ® fe3 a3 3 "2 p,2 PB Ti 2 o C w C sa 3§ CB ^3 X ® « fl jxS 3a go ,d’i 53^ ® g ® o 3 a ■£ ® ® 35a § * fl c3 (H Go © C3 fl ®,4> fl 03 fl •—I O'""1 •*-' O'—! o k.,G fl ,0 k..O SSSSSSa £ow_^o ’ Hiss 2 2 a r Table 11. — Results of baking tests with flour from fumigated wheat 32 BULLETIN 1313, U. S. DEPARTMENT OE AGRICULTURE Ethyl acetate ($5 per cent) FUMIGATION AGAINST GRAIN WEEVILS aa •S' 0 a 03 3 .a 1 .a > PH o g 6 o Is a o o a & . . . .a . t- ^ O OOOojOOfl o P pPP^P'gg p O fl O £ to ; a . ^ . o 'o >, a ^ ^ 03 • c3 t- h o) ® to ^ i CO O to 00 CO o N N *H N C5 00 00 05 05 c3 i a <3 M) | g fe d c3 U a O u u 0 o3 b£f^ o3 > M 1^2^ c3 !PO> P 00 OtOOOCO O 00 »o 01 o cd r-i td td 00 00 00 00 O CO 00 00 00 00 to O CO o OO ^ M 05 00 O to to O to o o oo ci oi oo c5 o 00 00 00 00 00 00 00 00 00 00 00 00 00 00 —i CO O* O 00 O 05 05 C5 00 to CO (M OOOOiCl o rji lO to to tototototototo to to o o o o o o o o o o o 05 CO 05 i— ( 00 CO O 0500000 © os' HNNNH CS cs ^ . d n to »o to »o to r-< O C5 N rH 03 05 CO CS 05 05 IO 05 oi cd c4 -? o o £a;c3a>Jr!■& >>£ &•“ b’D 5S5S-SS5S WQWOWOWO 34 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE EFFECT OF FUMIGATION ON MILLED AND BAKED PRODUCTS The results of the milling and baking tests on wheat and rye treated with various fumigants are given in Tables 10, 11, 12, and 13. Table 12. — Results of milling tests with rye exposed to fumigants Sample No. Fumigant Concen- tration of fumigant Odor of fumigant on grain 10073 Untreated _ Lbs. per 1,000 cu.ft. Natural. } Do. 10074 /Methyl formate _ 4. 0 \Ethyl bromide 11. 2 10075 /Ethylene bromide 8. 0 } Do. "ICarbon tetrachloride 12. 0 10076 /Ethyl formate 8. 8 1 Do. \Ethyl bromide 18.5 10077 Untreated Do. 10078 10079 do Do. j-None. /Ethylene bromide __ 8. 8 \Carbon tetrachloride. . ... 12. 0 10080 Untreated Natural. } Do. 10081 6. 4 \Ethyl bromide 13. 6 [Ethyl bromide 4. 5 10082 •[Carbon disulphide 2. 2 ^ Do. ICarbon tetrachloride 10.0 Table 13. — Results of baking tests with flour from fumigated rye 1 Sample No. Fumigant Concen- tration of fumigant Absorp- tion Volume of loaf Weight of loaf Color of loaf Texture of loaf 10073 10074 Untreated Lbs. per lfiOOcu.ft. Per cent 51.5 | 61.8 Cc. 1,630 1,720 Grams 462 485 73.0 74.0 90.8 /Methyl formate 4. 00 11. 20 90.3 10075 /Ethylene bromide 8. 00 } 51.5 1,800 488 75.5 91.8 \Carbon tetrachloride 12.00 10076 /Ethyl formate 8. 80 | 51.8 1,740 486 73.0 90.3 \E thy 1 bromide 18. 50 10077 51. 8 1,710 1,730 483 70. 8 89. 3 10078 10079 51.8 | 51.8 483 485 71.5 75.0 89.5 91.0 8.00 1,730 12.00 10080 10081 52.1 } 52.1 1,760 1, 710 484 479 72.5 73.0 90.5 89.5 6. 40 13. 60 [Ethylene bromide. 4. 50 10082 •j Carbon disulphide _ I Carbon tetrachloride . . 2. 20 10. 00 [ 51.8 1, 730 485 70.8 87.5 iln all the tests the crumb of the loaf was gray. No odor of the fumigant was detected in samples 10074, 10076, 10081, or 10082. A slight odor of the fumigant in the dough, but none in the loaf, either hot or cold, was detected in samples 10075 and 10079. Tetrachloro ethane and monochlorobenzene gave a disagreeable odor to the flour and bran and other by-products. This disagreeable odor in the flour was carried through to the finished loaf. Ethyl bromide, ethylene bromide, carbon tetrachloride, and ethyl formate left very little or no odor in the flour. If present in the flour, it was completely volatilized dining baking, giving a loaf free from foreign odor. The commercial grades of ethyl acetate, both the 85 per cent and the 99 per cent, left a noticeable odor in the grain, bran, shorts, and flour; this odor appeared even in the hot loaf. The FUMIGATION AGAINST GRAIN WEEVILS. 35 “old process” and “new process” grades of ethyl acetate completely volatilized from the gram, so that no odor was noticeable in the process of milling, and the loaf, hot or cold, had only the natural odor and flavor (samples 11024 and 11025 in Tables 10 and 11). The suitability of a sample of ethyl acetate for fumigating grain should be determined by the following test: Wet a sheet of filter paper (11-centimeter diameter is a convenient size) with the ethyl acetate, and allow it to evaporate, noting the odor from time to time. No foreign odor should be present, and the liquid should volatilize completely without leaving any odor. The carbon tetra- chloride to be mixed with the ethyl acetate should be similarly tested, and should likewise be free from odoriferous constituents of low volatility. The presence of sulphur compounds in carbon tetra- chloride is particularly objectionable because they give a garlicky odor to the fumigated grain. Ethyl acetate and carbon tetrachloride of a grade which satisfies this test are now commercially available in large quantities at a price but slightly higher than that asked for the commercial grades. ADDITIONAL FUMIGATION TESTS WITH ETHYL ACETATE AND CARBON TETRACHLORIDE In August and September, 1924, car fumigation tests were made with a mixture of ethyl acetate and carbon tetrachloride combined in a ratio that would make a noninflammable product at ordinary temperatures. The results are shown in Table 14. Table 14. — Results of fumigation tests on weevils in wheat in box cars, using a mixture of 40 volumes of ethyl acetate with 60 volumes of carbon tetrachloride Date Mean tem- pera- ture Rela- tive- humid- ity (noon) Con- cen- tration Aver- age mor- tality 1 Insects killed at different levels after exposure for 24 hours Insects Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Level 7 Level 8 Level 9 Level 10 in grain i Lbs. per 1,000 Per Per Per Per Per Per Per Per Per Per 1924 °F. Per cent cu.ft. Per cent cent cent cent cent cent cent cent cent cent cent Per cent Aug. 26 78 68 30 80 100 100 50 50 50 100 50 100 100 100 100 26 78 68 40 96 100 100 100 100 100 80 80 100 100 100 (3) 27 78 41 40 99 100 100 100 100 100 100 90 100 100 100 100 27 78 41 40 99 100 100 100 100 100 90 100 100 100 100 100 27 78 41 40 100 100 100 100 100 100 100 100 100 100 100 99 28 80 61 50 100 100 100 100 100 100 100 100 100 100 100 100 Sept. 3 66 63 40 100 100 100 100 100 100 100 100 100 100 100 100 3 66 53 40 100 100 100 100 100 100 100 100 100 100 100 100 4 66 45 40 4 86 100 100 100 100 100 80 90 90 80 20 100 6 60 41 40 98 100 100 100 100 100 100 100 80 100 100 (5) 6 60 41 40 97 90 100 90 100 100 100 100 100 100 90 0) 2 78 51 40 No weevils; sampled for milling and baking tests. 2 78 51 40 Do. 2 78 51 40 Do. 2 78 51 40 Do. 1 For insects exposed in pill boxes. 1 The grain in all but six of the cars fumigated was moderately infested with weevils. The percent- ages given in this column represent the kill of this infestation. 1 A few alive in ends of car. 4 Fumigation interrupted when halfway through for 10 or 15 minutes, during which time the car door remained open. '•> No weevils. 36 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE 05 a a -a „ fl s* ,fs 3S c3 <§ © © o o U. U o o caa c3 S 33 a a gJ c3 . . 6) fc& P P ^5 OP p 1 11 5&S ■5 3 !a ? ra fs n" a lu a o3 a So £ OSD & b •- s-i 3 i~< 3 ora o O 3 O -LTfl 11 1 >•*' o ^ o w . t> w 'OP 'd'tf ©Q © © 3 33 G5 a § 22 22 22 A A * * rP ►» •f a (d 03 © L L ^ . fcX) t>£) © q3 Ul J-. .42 p p p 3 oo O'T 3253 53 5> IT *2” IT ^ ^ c3 c5 q sR "B *3 a a a is & o3 c3 42 t-i u, t- o t£) bo ti) C/3 T3 T3 T3 © © © © £S ££ ££ E OT g M 6 3 t» 3 >, -O 3 (» _ 42 j-i 42 j-4 42 ^ » o TO © 03 © z> z> z> 3 £ 3 >, 3 >, -§ 42 *-< t2 »- 42 *-< 1 C3 © 03 © © © £> z> z> O S'd *cs & c, £ Co o W ' “ t- o 3T3 TO >J ■g >, 3-3 ' ' ® I ® ^ ‘ :> m> £ „ 1 '-t! o3 O OWiW ^ O P'P 53 >, "§ >> PP5 • • j_i p; j-< 42 1 © H © co > m> % r-i ^ ^ _ CO ' ‘ © O © O j- O O h *.3> 5? 53 £53 1 rt- ‘t- 3 ! ! ° O £ i “ a^ 's 53 02W 1? o © o © o © ©_< ©42 ©42 © 42 42*-* tt«! pq^ pq<< <1 m<< pq<< pq<< pq«< pq«< *0 CO TJ« TT* CO CO CO Tj* O o rt* — CO cO CO 03 CO co CO CO NM NiO COS 0003 Or O ip COCO *0*0 io *o COCO *0*0 coco coo co coco coco coco coco coco coco *0 Tf Tj* *0 1 Odor of fumigant in flour so slight as to be detected only by comparison with normal flour. The odor left by this fumigant is almost indistinguishable from that present in slightly fermented grain and is not objectionable from the standpoint of the baker because no odor remains in the bread baked from this flour. 2 This car of wheat was fumigated with the mixture at the rate of 40 pounds per 1,000 cubic feet; two days later the same wheat was again fumigated at the rate of 50 pounds per 1,000 cubic feet. 3 This wheat contained some seed of garlic, the odor of which was so pronounced that it masked all other odors. and 60 volumes of carbon tetrachloride FUMIGATION AGAINST GRAIN WEEVILS 37 o s o ■5 re £ -C 8 CC W r- 1 This car of wheat was fumigated with the mixture at the rate of 40 pounds per 1,000 cubic feet; 2 days later the same wheat was again fumigated at the rate of 50 pounds per 1,000 cubic feet. 38 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE With the exception of four cars fumigated for milling and baking tests, the rice weevil (S. oryza ) was the test insect used in the pill boxes. Furthermore, the grain in all but two of the cars was moderately infested with weevils. The results given in the last column in Table 14 show the percentage kill of this natural infestation. The results of the car fumigation reported in Table 14 confirm those reported in Table 8. With the exception of one case, where fumigation was inter- rupted, all fumigations with a mixture of 40 per cent by volume of ethyl acetate and 60 per cent by volume of carbon tetrachloride, in which not less than 40 pounds of the mixture per 1,000 cubic feet of inclosed space was used, gave satisfactory kills. The results of the milling tests with the wheat fumigated in these box-car experiments are shown in Table 15; the results of the baking tests with flour made from this wheat are shown in Table 16. The data indicate that both the flavor and odor of bread baked from flour made from wheat fumigated in box cars with a mixture of 40 volumes of ethyl acetate and 60 volumes of carbon tetrachloride, at the rate of 40 pounds per 1,000 cubic feet, are normal. EFFECT OF ETHYL ACETATE-CARBON TETRACHLORIDE FUMIGATION ON GERMINATION OF SEEDS The seeds of wheat, barley, rye, winter oats, and corn in tightly closed bell jars were exposed to the vapor of ethyl acetate and carbon tetrachloride at a concentration of 40 pounds per 1,000 cubic feet for 24 hours. The percentage germination before and after fumiga- tion is shown in Table 17. Table 17. — Effect of ethyl acetate-carbon tetrachloride fumigation on germination of seeds Seed Germination Seed Germination Before fumiga- tion After fumiga- tion Before fumiga- tion After fumiga- tion Per cent 83 97 89 Per cent 85 98 90 Per cent 95 97 Per cent 93 94 Rye..'. The results in Table 17 show that exposure at 31° C. in a tight container for 24 hours to the fumes of a 40-60 mixture (by volume) of ethyl acetate and carbon tetrachloride at the rate of 40 pounds per 1,000 cubic feet does not injure the germinating power of wheat, barley, rye, winter oats, and com. The grain would not be subjected to as severe a test in ordinary box cars. SUMMARY The action of more than 100 organic compounds on weevils was tested under conditions permitting a control of the factors of con- centration, time, and humidity, and with observations of the tem- perature. The following 30 compounds were more toxic to the rice weevil (JS. oryza L.) than carbon disulphide: Two out of 8 bromides tested; FUMIGATION AGAINST GRAIN WEEVILS 39 3 out of 15 chlorides tested; the only iodide tested; 1 alcohol out of 8 alcohols and phenols tested; 3 out of 4 aldehydes tested; 2 out of 4 ketones tested; both of the chlorine-substituted ketones tested; 1 of the 2 chlorohydrins tested; 5 out of 9 esters tested; 4 out of 12 sulphur compounds tested; 1 nitrile out of 5 nitriles and isonitriles tested; both of the nitrites tested; 2 out of 8 amines tested; and pyridine. The relative toxicity of the different classes of compounds can not be given because the low volatility of several of those tested gave only very low vapor concentrations. As a class, the hydro- carbons showed the lowest insecticidal efficiency, not one of the eight tested equaling carbon disulphide in fumigating power. The most effective fumigant in the glass-jar tests was epichlorohydrin, which killed the rice weevil at a concentration of 0.09 per cent, equivalent * to 0.23 pound per 1,000 cubic feet. It was, however, an unsatisfactory fumigant in the presence of grain. There is no constant relationship between the boiling points and the lethal concentrations of the compounds killing 100 per cent of the rice weevils after exposure for 24 hours. A much greater concentration of fumigant is required to kill weevils in wheat than to kill those exposed directly to the vapors in glass jars. A still higher concentration is necessary to kill weevils in wheat in box cars. Ethyl formate and ethyl acetate were the only promising fumigants for grain in box cars. The acetate, however, costs only about one-third as much as the formate. Odoriferous constituents of low volatility from commercial grades of ethyl acetate (both the 85 per cent and the 99 per cent grades) are carried through from the fumigated wheat to the flour, and even to the bread baked from it. A pure grade of ethyl acetate, however, leaves practically no odor in the fumigated grain or in the bran or shorts made from the grain, and none in the flour or in the bread baked from the flour. The insecticidal efficiency of ethyl acetate under practical fumigat- ing conditions is increased by the addition of carbon tetrachloride. The most effective fumigant, other than carbon disulphide, against weevils in wheat, in grain cars, under practical fumigating conditions, is a mixture of about 40 volumes of ethyl acetate and about 60 volumes of carbon tetrachloride. It is noninflammable at ordinary temperatures. The proper dosage of this mixture for fumigating box cars is about 45 pounds per 1,000 cubic feet. Both the ethyl acetate and the carbon tetrachloride must be tested to make sure that they are free from odoriferous constituents of low volatility before they are used in grain fumigating. LITERATURE CITED (1) Altson, A. M. Beetles damaging seasoned timber — IV. Methods of treatment. In Timber Trades Journal (1922), 51:1170-1. (2) Bertrand, Gabriel, and Rosenblatt, M. Action toxique compare de quelques substances volatiles sur divers insectes. In Compt. rend. (1919), 165:911-13. (3) Burmeister, Hermann. A manual of entomology, translated bv W. E. Shuckard, London (1836), p. 391. (4) Carteret, M., and Carteret, G. Sur l’alt^ration des farines et c4r6ales par S02. In Bull. soc. chim. (1909;, series 4, vol. 5, p. 270-2. 40 BULLETIN 1313, U. S. DEPARTMENT OF AGRICULTURE (5) Davis, John J. Miscellaneous soil insecticide tests. In Soil Science (1920), 10:61-72. (6) Foreman, F. W., and Graham-Smith, G. S. Investigations on the prevention of nuisances arising from flies and putrefaction. In J. Hygiene (1917), 16: 109-226. (7) GuArin, P., and Lormand, Ch. Action du chlore et de diverses vapeurs sur les v£g6taux. In Compt. rend. (1920), 170: 401-3. (8) Harcourt, It. Effect of mill fumigants upon flour. In Ontario Agr. Coll. Expt. Farm 36th Ann. Rept., 1910 (Toronto, 1911), pp. 87-92. (9) Holt, Joseph J. H. The cockroach: Its destruction and dispersal. A comparison of insecticides and methods. In Lancet (1916), 190: 1136-7. (10) Jewson, Sibyl T., and Tattersfield, F. The infestation of fungus cultures by mites: Its nature and control, together with some remarks on the toxic properties of pyridine. In Ann. app. biol. (1922), 5:213-40. (11) Lefroy, H. M. Insecticides. In Ann. app. biol. (1915), 1:280-98. (12) Lloyd, L. The control of the greenhouse white fly ( Asterochiton vo.porariorum ) with notes on its biology. In Ann. app. biol. (1922), 5:1-32. (13) McClintock, C. T., Hamilton, H. C., and Lowe, F. B. A further contribution to our knowledge of insecticides. Fumigants. In J. Am. Pub. Health Assoc. (1911), 1: 227-38. (14) Moore, William. Toxicity of various benzene derivatives to insects. In J. Agr. Research (1917), 5:371-81. (15) . Volatility of organic compounds as an index of the toxicity of their vapors to insects. In J. Agr. Research (1917), 15:365-371. (16) . Methods of control of the clothes louse [Pediculus humanus ( vestimenti)\ . In J. Lab. Clin. Med. (1918), 5:261-8. (17) and Graham, S. A. Toxicity of volatile organic compounds to insect eggs. In J. Agr. Research (1918), 12: 579-87. (18) Neifert, I. E., and Garrison, G. L. Experiments on the toxic action of certain gases on insects, seeds, and fungi. U. S. Dept. Agr. Bui. 893 (1920), 16 pp. (19) Parker, T., and Long, A. W. A laboratory note on the control of Trogoderma khapra. Bureau of Bio-Technology Bui. 4 (1921), pp. 102-04. (20) Phelps, E. B., and Stevenson, A. F. Experimental studies with muscicides and other fly-destroying agencies. U. S. Pub. Health Service, Hyg. Lab. Bui. 108 (1916), pp. 5-30. (21) Richardson, Charles H., and Smith, C. R. Studies on contact insecticides. U. S. Dept. Agr. Bui. 1160 (1923), 15 pp. (22) Russell, E. J. The partial sterilization of soils. In J. Roy. Hort. Soc. (1920), 45: 237-46. (23) Speyer, E. R. Report of the entomologist. Experimental and Research Station, Nursery and Market Garden Industries’ Development Society, Ltd., 8th Ann. Rept., 1922. Cheshunt, England (1923), pp. 45-57. (24) Tattersfield, F., and Roberts, A. W. R. The influence of chemical constitution on the toxicity of organic com- pounds to wireworms. In J. Agr. Sci. (1920), 15:199-232. (25) Titschack, E. Beitrage zu einer Monographie der Kleidermotte, Tineola biselliella Hum. In Zeit. tech. Biol. (1922), 15:1-168. (26) Trillat and Legendre, J. Etude sur la toxicity des vapeurs de quelques substances chimiques sur les moustiques. In Hygiene g6n6rale et appliqu^e (1909), 4:542-546. \ ORGANIZATION OF THE UNITED STATES DEPARTMENT OF AGRICULTURE Jannary 1, 1923 Secretary of Agriculture Assistant Secretary Director of Scientific Work Director of Regulatory Work Director of Extension Work Solicitor Weather Bureau Bureau of Agricultural Economics Bureau of Animal Industry Bureau of Plant Industry Forest Service Bureau of Chemistry Bureau of Soils Bureau of Entomology Bureau of Biological Survey Bureau of Public Roads Bureau of Home Economics Bureau of Dairying Fixed Nitrogen Research Laboratory Office of Experiment Stations Office of Cooperative Extension Work — Office of Publications Library Federal Horticultural Board Insecticide and Fungicide Board Packers and Stockyards Administration Grain Futures Administration Howard M. Gore. E. D. Ball. Walter G. Campbell. C. W. Warburton. E. W. Williams. Charles F. Marvin, Chief. Henry C. Taylor, Chief. John R. Mohler, Chief. William A. Taylor, Chief. W. B. Greeley, Chief. C. A. Browne, Chief. Milton Whitney, Chief. L. O. Howard, Chief. E. W. Nelson, Chief. Thomas H. MacDonald, Chief. Louise Stanley, Chief. C. W. Larson, Chief. F. G. Cottrell, Director. E. W. Allen, Chief. C. B. Smith, Chief. L. J. Haynes, Director. Claribel R. Barnett, Librarian. C. L. Marlatt, Chairman. J. K. Haywood, Chairman. Chester Morrill, Assistant to the Secretary. This bulletin is a contribution from Bureau of Chemistry C. A. Browne, Chief. Miscellaneous Division J. K. Haywood, Chief. Bureau of Entomology L. O. Howard, Chief. • Stored Product Insect Investigations E. A. Back, In charge. ADDITIONAL COPIES or THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY V Tr~ ~ - UNITED STATES DEPARTMENT OF AGRICULTURE Washington, D. C. ▼ April, 1925 BEHAVIOR OF COTTON PLANTED AT DIFFERENT DATES IN WEEVIL- CONTROL EXPERIMENTS IN TEXAS AND SOUTH CAROLINA By W. W. BALLARD, Senior Scientific Aid, and D. M. SIMPSON, Assistant Agronomist Office of Crop Acclimatization and Adaptation Investigations Bureau of Plant Industry CONTENTS Page Introduction 1 Soil, Climatic, and Weevil Conditions at San Antonio, Tex 2 Comparison of Successive Adjacent Plantings at San Antonio 3 Yields from Successive Plantings at San Antonio 19 Percentage of 5-Lock Bolls on Early and Late Plantings at San Antonio . 22 A Separate Late Planting at San Antonio 23 Yields from Cotton Experiments at San Antonio 28 Soil, Climatic, and Weevil Conditions at Charleston, S. C 29 Comparison of Successive Adjacent Plantings at Charleston 30 Yields from Successive Plantings at Charleston 38 Adverse Conditions at Gainesville, Fla 41 Summary 41 WASHINGTON GOVERNMENT PRINTING OFFICE 1925 UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1320 Washington, D. C. V April, 1925 BEHAVIOR OF COTTON PLANTED AT DIFFERENT DATES IN WEEVIL- CONTROL EXPERIMENTS IN TEXAS AND SOUTH CAROLINA Bv W. W. Ballard, Senior Scientific Aid , and D. M. Simpson, Assistant Agron- omist, Office of Crop Acclimatization and Adaptation Investigations, Bureau of Plant Industry CONTENTS Page Introduction 1 Soil, climatic, and weevil conditions at San An- tonio, Tex. - 2 Comparison of successive adjacent plantings at San Antonio 3 Yields from successive plantings at San An- tonio 19 Percentage of 5-lock bolls on early and late plantings at San Antonio 22 Page A separate late planting at San Antonio 23 Yields from cotton experiments at San Antonio. 28 Soil, climatic, and weevil conditions at Charles- ton, S. C 29 Comparison of successive adjacent plantings at Charleston 1 30 Yields from successive plantings at Charleston. 38 Adverse conditions at Gainesville, Fla 41 Summary 41 INTRODUCTION More information is needed on the growth and fruiting habits of early and late planted cotton in relation to cultural control of the boll weevil. In the season of 1923 comparisons were made of the behavior of early and late plantings in Texas and South Carolina, and differences were shown in the rates of growth and the fruiting habits of the plants. A more rapid formation of nodes during the seedling stage of the plants was found to occur in the later plantings, resulting in a shorter interval between the date of planting and the appearance of the first floral bud. The fruiting capacity of late-planted cotton was found to equal and in some cases to exceed that of the early-planted cotton. The large number of floral buds produced in later plantings was due to the fact that more nodes were produced on the lower fruiting branches. Also, slightly larger numbers of flowers were recorded on the late-planted cotton, although the early plantings produced a larger number of flowers during the first part of the flowering period. The experiments were made in three places — San Antonio, Tex., Charleston, S. C., and Gainesville, Fla. The object of having similar tests in three widely separated parts of the Cotton Belt was to secure comparative data of plant development under different soil and climatic conditions. The experiments consisted of side-by- side comparisons of cotton planted on four different dates. An 20718°— 25 1 1 2 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE interval of 10 days between plantings was used in South Carolina and Florida and 7 days in Texas. As these experiments were conducted under boll-weevil conditions, measures were taken to protect the early plantings from infestation by overwintered weevils. Uncontrolled infestation in the early plantings would have caused excessive infestation in the adjacent later plantings and would have prevented comparable data being obtained from the different plantings. The method of weevil con- trol was by removal of squares and the application of poison after most of the weevils had emerged from hibernation. At San Antonio, Tex., it was only necessary to apply pois n to the two later plantings, those of May 5 and May 12, as these plant- ings had not reached the fruiting stage at the time squares were removed from the early plantings. A separate late planting was made at San Antonio on May 12 in order to compare the development and fruiting habits of cotton in rows that were left unthinned with rows that were thinned to two plants in a hill with hills 12 inches apart. This comparison showed that plants left in hills had a larger individual fruiting capacity than the unthinned plants, due to the production of more nodes on the fruiting branches. But this difference in the fruiting capacity of individual plants was counterbalanced by the greater number of plants in the unthinned rows. More flowers were recorded from the unthinned cotton, and a marked difference in the rate of flowering occurred during the first half of the flowering period, the unthinned rows producing flowers at almost twice the rate of the rows that were thinned. No trace of wTeevil infestation -was found in this separate late- planted field until July 8, after the beginning of the flowering period. Thus, it appeared that the planting of May 12 had been late enough to avoid any infestation from overwintered weevils in the season of 1923 at San Antonio. SOIL, CLIMATIC, AND WEEVIL CONDITIONS AT SAN ANTONIO, TEX. The United States San Antonio Field Station is located about 5 miles south of the city. The soil on the farm is typical of a large part of the cultivated land in that region and is technically described as Houston clay loam. It is of high natural fertility and is fairly retentive of moisture. The subsoil is a coarse gravel, which affords good drainage but limits the water-storage capacity of the soil. The annual precipitation is variable, ranging from 13 to nearly 40 inches. The average annual rainfall over a period of 15 years is about 26 inches. Although the precipitation is usually sufficient for cotton production, the distribution is very irregular. Periods of excessive rainfall are frequently followed by protracted periods of hot dry weather which deplete the soil moisture and cause serious injury to crops. The precipitation during the season of 1923, from January 1 to October 1, was 23.47 inches, which was about 4 inches in excess of the 15-year average for the same period. The heaviest monthly rainfall, 6 inches, was recorded in February. Rains occurring during the winter months are often an important factor in this section, as stored moisture in the soil ma}^ enable the plants to continue growth COTTON IN WEEVIL-CONTROL EXPERIMENTS 3 even when surface moisture becomes deficient. June was very dry, with only 0.55 inch of rain. Precipitation during the growing period of cotton was fairly well distributed, but effective rains fell on only six days between May 30 and September 7. Maximum temperatures during June, July, and August were comparatively high, and the long periods of hot weather between the few good rains resulted in droughty conditions in spite of the appar- ently adequate precipitation. The average monthly maximum and minimum temperatures and the monthly precipitation for the season of 1923 are given in Table 1. Table 1. — Average minimum and maximum temperatures and monthly precipita- tion at San Antonio, Tex., from April to September, 1923 Items of comparison Jan. to Mar. Apr. May June July Aug. Sept. Average temperature (° F.): 59.3 64.5 72.4 71.2 72.4 70. 7 77.6 89.0 95.4 95.3 97.8 89.4 Precipitation (inches) : 1923- 9.27 2. 93 1.67 . 55 3. 77 2.50 3. 02 3. 98 3. 11 3. 09 1.88 2. 12 1.87 2.57 Favorable conditions for the hibernation of boll weevils in the San Antonio district are afforded by scattered areas of undeveloped land covered with mesquite and huisache trees and by large fields of Johnson grass. A heavy infestation from overwintered wTeevils usually occurs in cotton planted in this section. The first new7" genera- tion of weevils usually appears before the first of July. Periods of dry weather often restrict the growth of the plants, and the shed squares are exposed to direct sunlight, so that a natural control of the weevils may result from destruction of the larval and pupal stages in the fallen squares. When heavy rainfalls occur during the early stages of plant growth, excessive vegetative develop- ment of the plants results, so that the weevils have more protection even though there is dry weather later in the season. There is slight possibility of effective natural control when the plants grow large and the lanes between rows are shaded so as to protect the infested squares. In the season of 1923 weevils were first found on May 28. Squares were very small at that time, but by the first week in June numerous punctured squares could be seen in all cotton which had reached the fruiting stage. Dry weather throughout the greater part of May and June retarded the growth of the plants. This small size kept the lanes open between the rows and afforded favorable conditions for natural control of weevils by exposing the fallen squares to direct sunlight. Nevertheless, the weevil infestation increased slowly, until by the first of August there wrere enough weevils to destroy practically all of the squares. COMPARISON OF SUCCESSIVE ADJACENT PLANTINGS AT SAN ANTONIO At San Antonio the successive plantings were made on April 19 and 28 and May 5 and 12. The Lone Star variety was used, the seed having been grown on the experiment farm in 1922. The rows were 4 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE 200 feet in length and were spaced 4.1 feet apart. The location of this test with relation to other cotton plantings on the experiment farm is shown in Figure 1, and the planting diagram of the test is shown in Figure 2. Conditions were favorable for the germination of the seed in the first three plantings, and good stands were obtained. Dry weather, following rains which occurred during the last week in April, dried out the soil to a considerable depth, and a stand of about 30 per cent resulted from the planting of May 12, the remainder of the seed lying in the ground until germinated by a rain on May 30. The seedlings produced from the seed that germinated first grew rapidly and pro- duced vigorous plants. The moisture supply in the surface soil was inadequate to support growth of the seeds that germinated late, and most of the resulting seedlings died or remained stunted throughout the season. Fig. 1. — Diagram of part of the United States San Antonio Field Station, showing the location of successive plantings and the separate late-planting test of cotton with relation to other cotton plantings The first three plantings were thinned to two plants in a hill with the hills 12 inches apart. Thinning was done by hand in order to obtain as regular spacing as possible. The fourth planting was not thinned, as the poor stand rendered this unnecessary. The April 19, April 28, and May 5 plantings were thinned on May 25, June 4, and June 8, respectively. At the time of thinning, the plants in the first two plantings had from five to seven true leaves and averaged about 20 centimeters in height. The plants in the third planting were slightly larger when thinning was done, averaging 22.5 centimeters in height and having from six to eight true leaves. GROWTH RATE OF SEEDLINGS In order to compare the rate of development during the seedling stage, the height and number of nodes on 10 representative plants in each planting were recorded when the first floral buds, or squares, appeared. The plants in each planting were examined on June 1 and COTTON IN WEEVIL-CONTROL EXPERIMENTS 5 ^ -C‘7'/OA/s9 ^E A? AA/.9 Fig. 2.- Diagram of successive plantings of cotton at San Antonio, Tex. Table 2. — Average development of 10 cotton plants at the beginning of the fruiting stage as grown in successive plantings at San Anionio on four different dates Date planted, 1923 Squares first recorded Number of days from planting Height of plants (centi- meters) Number of nodes on main stalk Average number of squares Apr. 19 43 21. 5 0.5 2.2 Apr. 28 41 23.4 7. 1 2.8 May 5 ...do 34 22.0 0.2 1.9 May 12 June 15 34 24.4 5.9 1.7 1 All records of the number of nodes on the main stalk are exclusive of the cotyledon node. 6 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE Table 3. — Relation of temperature to the development of cotton plants as grown in successive plantings at San Antonio on four different dates [The temperatures shown (° F.) cover the period from the date of seed germination to the first appearance of squares] Date planted, 1923 Date of germi- nation Number of days to squaring Number Average number Average tempera- ture formed of days per node Mini- mum Maxi- mum Apr. 19_ ___ Apr. 25 May 4 May 12 May 20 37 6. 5 5. 7 64.4 88. 5 Apr. 28 35 7. 1 4. 9 66. 5 90 8 27 6. 2 4.4 68. 6 92 3 May 12 26 5.9 4.4 70.7 94.0 It will be noted that the interval between the development of suc- cessive nodes decreases when higher temperatures occur during the period from germination to the appearance of squares. The average rate of nodal development during the seedling stage of the April 19 planting was 5.7 days. During this period the average minimum temperature was 64.4° and the average maximum 88.5° F. The May 12 planting averaged 4.4 days per node during the seedling stage of the plants, the average minimum and maximum temperatures having been 70.7° and 94° F., respectively. DAMAGE BY OVERWINTERED WEEVILS The extent of weevil infestation in the successive plantings of cot- ton was determined on June 5 by recording the numbers and per- centages of squares that had been attacked by weevils. On June 5 squares were present only on the April 19 planting. Even on this there were only a few large squares, and it was necessary to examine about 200 plants in order to find. 100 squares that were regarded as over 10 days old. Records of growth of squares have indicated that a period of about 10 days elapses between the time when a square is first visible and the time when it has reached sufficient size to harbor a weevil larva.2 • In block 1 of the April 19 planting, which was located on the south side of the field, 45 per cent of the squares were punctured by weevils. Block 2, in the center of the field, had 20 per cent of the squares punctured, and block 3, on the north side of the field, had 15 per cent of punctured squares. I'he heavier infestation in block 1 may have been due to the fact that the field adjoined a Johnson grass pasture on the south. A large part of the weevil emergence probably occurred from this pasture, and a large number of the weevils would doubtless remain in the first cotton where squares were found. STRIPPING OF FLORAL BUDS OR “SQUARES” Squares were removed on June 12 from the first and second plant- ings, and the entire field was poisoned with calcium arsenate. At that time the squares on the third and fourth plantings were below the size for stripping, so that it was necessary only to apply poison. 2 Martin, R. D., W. W. Ballard, and D. M. Simpson. Growth of fruiting parts in cotton plants. In Jour. Agr. Research, v. 25, p. 202. 1923. COTTON IN WEEVIL-CONTROL EXPERIMENTS 7 No squares were removed that were less than 10 days old, at which time the involucre of the square was about three-eighths of an inch in length. It was found to be impossible to remove squares smaller than this without breaking the tips of the fruiting branches or injuring the terminal buds of the plants. Records were obtained of the number of squares picked from four rows in the April 19 and April 28 plantings. The number of plants to the row and the number of squares removed from each row are given in Table 4. Table 4. — Number of squares removed on June 12 from cotton plants of the first and second planiings in rows 200 feet long at San Antonio Number of plants to the Number Average number Number of plants to the Number Average number Date planted, 1923 of squares of squares Date planted, 1923 of squares of squares removed to the removed to the plant plant ( 392 787 2.0 300 128 0.4 Apr. 19 1 375 403 803 823 2. 1 2.0 Apr. 28 I 261 285 146 114 .6 .4 1 387 734 1.9 1 334 193 . 6 An average of two squares removed from each plant corresponds to the amount of stripping reported in experiments with the stripping method in Florida.3 A difference of nine days in the planting date of the first and second plantings resulted in only one-fourth as many squares being removed from the second planting as from the first. In plantings which were deferred until May 5 and May 12 there was no necessity of square removal to avoid infestation by overwintered weevils. The average time required for finding and removing squares was 48 minutes per row in the April 19 planting and 22 minutes per row in the planting of April 28, the length of rows being 200 feet. On this basis the time required to strip an acre was estimated at 41 hours for the first planting and 19 hours for the second. At the rate of four days per acre, or even two days, the labor requirement for square stripping is considerable. Only 2 weevils were found in squares removed from the two blocks of the second planting, while 96 weevils were caught in the three blocks of the first planting. The fact that so few weevils were caught in the second planting is probably due to the smaller number of large squares on the later plants. Weevils which are feeding on the floral bud inside the involucre of large squares are more likely to be caught. A hundred squares from each row in the April 19 and April 28 plantings were examined, in order to determine the percentage of squares that had been punctured. The records of punctured squares in each block of the first planting gave the following average per- centages of infestation: Block 1, 23; block 2, 14; block 3, 12. The first block of the second planting had 9.7 per cent and the second block had 6.5 per cent of the squares punctured. Calcium arsenate in dry-dust form was applied to the entire field immediately following the removal of squares from the first two 3 Smith, O. D. A preliminary report upon an improved method of controlling the boll weevil. Fla. Agr. Exp. 3ta. Bui. 106, 72 p., illus. 1922. 8 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE plantings. A hand gun was used in dusting the plants, the poison being applied at the rate of about 8 pounds per acre. It remained on the plants until June 17, when most of it was washed off by a light rain. No further poisoning was attempted during the remainder of the season. A comparison of the development of plants in the four plantings was made on June 15, three days after squares had been removed and poison applied. Data on the height of plants, the number of nodes on the main stalk, the total number of squares that had been formed to June 15, and the actual number of squares on the plants are presented in Table 5. Table 5. — Development of cotton plants, showing the average number of squares formed, on each plant on June 15 as grown in successive plantings at San Antonio on four different dates Height (centi- meters) Number of nodes Average number of squares on plants Date planted, 1923 Formed to June 15 Remain- ing on June 15 28.6 11.2 10.6 5.9 Apr. 28 28.0 9.2 6.5 5.1 27.8 8.6 5.7 5.4 24.3 5.9 1. 7 1.7 The difference between the total number of squares that had been formed by June 15 and the number actually on the plants at that date represents the loss through square removal and shedding. Considerable shedding of very small squares occurred in the April 19 planting, practically all of the squares formed prior to June ,1 having been shed by June 8. WEEVIL INFESTATION AFTER STRIPPING SQUARES Although the young squares on the first three plantings had developed within a week after stripping and poisoning to a size which would render them susceptible to weevil injury, no indication of infestation was detected for another week, or until June 25. Three small areas of infested plants were observed on that date in widely separated parts of the field. The absence of •weevil infestation for a period of nearly two weeks after stripping and poisoning indicates that the control measures had practically exterminated the weevils present in the field on June 12. It is probable that the slight infestation which was first noted on June 25 resulted from weevils that had emerged from early-punc- tured squares. Some of the squares on the first planting which had been punctured during the first week in June were shed before the plants were stripped. Some of the squares which had been shed previous to stripping might have been missed by laborers. Several scattered points of infestation appeared within a few days after the first trace of weevil damage was noted, most of these points occurring in the first planting. Infestation increased slowly, and by the middle of July evidence of weevil injury could be found through- out the field. COTTON IN WEEVIL-CONTROL EXPERIMENTS 9 Although the field became reinfested with weevils after square removal and poisoning, the degree of infestation was much less than in untreated early-planted fields on other parts of the experiment farm. Possibly a more effective control might have been obtained if the measures had been applied earlier than June ' 12, as fewer infested squares would have been shed before the square-stripping operation. PLANT GROWTH DURING THE FRUITING PERIOD Records of the nodal growth of the cotton plants, the rate of flowering, and the extent of boll shedding were obtained from each of the four plantings. Comparisons of the rate of formation of internodes were obtained from 10 representative plants in each plant- ing, while the flower counts and boll sheds were recorded from 50-foot sections of rows. The records of plant development following the appearance of squares were obtained from diagrams of the same plants which were used in comparing the rate of growth during the seedling stages. These diagrams were made at weekly intervals throughout the period of fruiting of the plants, the final records having been made on August 11, after growth had practically ceased. A comparison of the rate of formation of internodes on plants in the successive plantings is shown in Table 6, which gives the average number of nodes on 10 plants of each planting on the date when squares were first recorded and 14, 28, and 56 days later. As the first squares were recorded on different dates, it should be noted that the nodal development of the plants in each planting is compared during similar stages of growth and not on the same dates. Table 6. — Development of nodes on the main stalk during the fruiting stage of cotton plants grown in successive plantings at San Antonio on four different dates Number of nodes at stated intervals after first appearance of squares Date of planting, 1923 Squares first recorded Aver- age 14-day interval 28-day interval 56-day interval Total Prob- Actual In- crease Actual In- crease Actual In- crease in- crease able error Apr. 19 6. 5 11.2 4.7 14.8 3.6 18. 6 3.8 12. 1 ±0. 15 Apr. 28 7.0 11. 5 4.5 14.9 3.4 18. 2 3.3 11.2 ± .22 May 5 6.2 10.9 4.7 14. 2 3. 3 16. 9 2.7 10. 7 ± .37 May 12 June 15 5.9 11.8 5.9 15. 1 3.3 18.9 3.8 13.0 ± .22 The average number of nodes produced on the main stalk during the first period of 14 days after the appearance of squares was 4.7, 4.5, and 4.7 nodes, respectively, on plants of the first three plantings. The planting of May 12, however, formed 5.9 nodes during this period. During the second period of 14 days the rate of formation of nodes was much slower, the increase in number of nodes having been 3.6, 3.4, 3.3, and 3.3, respectively. The final period comprised 28 days, but the increase in number of nodes was only 3.8, 3.3, 2.7, and 3.8, or about the same as during the preceding period of 14 days. The total number of nodes formed during the entire period of 56 davs after the appearance of the first squares was 12.1 nodes on plants 20718°— 25t 2 10 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE of the April 19 planting, 11.2 nodes on the April 28 planting, 10.7 nodes on the May 5 planting, and 13 nodes on the May 12 planting. In order to show the number of nodes on plants in the different plantings on the same dates, the weekly records of nodal develop- ment are graphically presented in Figure 3. These curves show that the nodal development of the first three plantings was nearly parallel throughout the period following the appearance of squares until August 11. The May 12 planting, however, produced nodes more rapidly throughout this period. From June 1 until August 11, a period of 71 days, the April 19 planting produced 13.1 nodes. From June 8 until August 11, a period of 64 days, the April 28 and May 5 plantings produced 11.1 and 11 nodes, respectively. From June 15 to August 11, a period of 57 days, the May 12 planting produced 13 nodes. The average number of days per node for each planting was 5.42, 5. 77, 5.82, and 4.38 days, respectively. Fig. 3. — Average number of nodes on the main stalks of cotton plants at weekly intervals following the appearance of squares at San Antonio, Tex. Each node produced on the main stalk after the beginning of the fruiting stage of the plants represents the formation of a fruiting branch. Thus, the May 12 planting produced the same number of fruiting branches in 15 days less time than was required by the April 19 planting. Figure 4 gives the average height of 10 plants in each planting at weekly intervals from June 1 until August 11. While the height of plants is a factor of less importance than the nodal development, it will be of interest as showing the relative size of plants in each planting. The more vigorous growth of plants in the later plantings is shown in these curves. On July 13 the plants of the May 12 plant- ing were the largest, with the May 5, April 28, and April 19 plantings following in the order named. Although the plants in the April 19 planting maintained a larger number of nodes throughout the season, after the middle of July they were the smallest. The difference in the size of representative plants in each planting is shown in Plates I and II. This tendency toward the development of larger plants in later plantings is to be expected, especially on the heavier types of Texas COTTON IN WEEVIL-CONTROL EXPERIMENTS 11 soils, when the moisture supply is plentiful. In the present experi- ment the late-planted cotton made considerably more vegetative growth than that planted early, in spite of drought conditions. The larger growth of the late-planted cotton resulted from the devel- opment of longer internodes on the main stalk and on the fruiting branches. Although the plants grew taller , they had fewer nodes, as shown in Figure 3. PRODUCTION OF FLORAL BUDS The fruiting capacity of a cotton plant is determined principally by the number of fruiting branches formed on the main stalk and the number of internodes formed on the fruiting branches. The plants in the successive plantings at San Antonio were small and formed few squares on secondary fruiting branches, and these are not included in comparisons of the fruiting capacity of plants. The average number of squares formed on 10 plants in each of the four plantings is given in Table 7. The numbers of squares were recorded at 2-week intervals throughout the fruiting stage of the plants, the final record on August 11 representing the total number of squares formed on the plants. Table 7. — Average number of squares recorded at 2-week intervals on cotton plants grown in successive plantings at San Antonio on four different dates Date planted, 1923 Apr. 19. A pr. 23. May 5. May 12 Number of squares recorded on — June 1 June 15 June 28 July 13 July 28 Aug. 11 2.2 10.6 21.7 26.6 31.9 35.3 0 6.5 18.6 26.4 33.3 38.7 0 5.7 17.6 23.1 25.8 28. 5 0 1.7 12.8 24.2 26.4 30.9 12 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE These records show that the plants in the April 19 and April 28 plantings produced a larger number of squares than those in the plantings of May 5 and May 12. The number of squares, however, increased more rapidly on the later plantings during the first half of the squaring period, as will be seen by a comparison of the number of squares on the plants by July 13. On this date the average num- ber of squares on plants of the May 12 planting was 24.2, as compared with 26.6 on plants of the April 19 planting, although this planting had reached the fruiting stage two weeks earlier than the Dlanting of May 12. The fact that the number of squares formed on the late plants by the middle of July was almost as great as the number on the early plants was largely due to a better development of the lower fruiting branches on plants of the later plantings. Table 8 gives the average number of internodes formed on fruiting branches. The numbers of internodes are determined from groups of fruiting branches, in order to simplify the presentation of the data. The first eight fruiting branches on plants of the May 5 and May 12 plantings developed more internodes than the first eight branches of the early plantings. The number of internodes on the upper fruiting branches was approxi- mately the same on the first and last plantings, while the upper fruit- ing branches on the second and third plantings were not so well de- veloped. This tendency toward the formation of more internodes on lower fruiting branches of late-planted cotton indicates that the growth of the plants was not retarded by the dry weather during June and the first part of July to so great an extent as with the early plants. This is shown in Plates I and II by the presence of squares on plants of the later plantings during the first part of August, after all squares had been shed from the plants in the early plantings. Table 8. — Average number of nodes on fruiting branches of cotton plants grown in successive plantings at San Antonio on four different dates Date planted, 1923 Branches 1 to 4 Branches 5 to 8 Branches 9 to 12 Branches 13 to 15 1.9 2.8 2.45 1. 27 1. 57 2. 87 2.07 .9 2. 17 3.07 2. 07 .8 May 12 2. 52 3. 62 2.92 1.0 The average number of internodes on the first eight fruiting branches on plants of the May 12 planting was 3.04, as compeared with 2.3 internodes on plants of the April 19 planting. An even more pronounced tendency toward the formation of a greater number of internodes on the lower fruiting branches of later plantings was shown in the experiments at Charleston. FLOWERING RECORDS OF EARLY AND LATE PLANTINGS Flower counts were obtained from four 50-foot sections of rows in the April 19, April 28, and May 5 plantings, and from two 50-foot sections of rows in the May 12 planting. A section of row was located on the east and west ends of rows in the first and second blocks of the April 19, April 28, and May 5 plantings. No flower counts were Bui. 1320, U. S. Dept, of Agriculture Plate I Size and Fruiting of Cotton Plants in Successive Plantings at San Antonio, Tex., on August 7. — I A, Planted April 19; li, planted April 28. (Compare with Plate II) Bui. 1320, U. S. Dept, of Agriculture Plate II Size and Fruiting of Cotton Plants in Successive Plantings at San Antonio, Tex., on August 7. — II A, Planted May 5; B, planted May 12. (Compare with Plate I) Note the larger size and production of vegetative branches on the plants in the May 12 planting, showing the need of closer spacing of plants in late-planted cotton. Also note the presence of squares and flowers on these plants, produced under drought conditions, which had checked growth in the early-planted cotton COTTON IN WEEVIL-CONTROL EXPERIMENTS 13 made in the west end of the May 12 planting because of the very poor stand. Flower counts were started on June 26 and were made daily until July 7. From that time until July 31 they were made every other day. The counts from each section of row in each planting are given in Table 9, with the total number of flowers counted each day in each planting shown graphically in Figure 5. The total number of flowers counted on the west end of the rows is usually smaller than the number on the east end. This was due to retarded plant growth in the west end of the rows caused by Johnson grass in that part of the field. — r — v — t/77/// A// £/A/£^7c/S £?//£///&*& £77/// A// SlZA/d^7c/ S J2V&T££//£A// *^S£/A/6/7c/ £zx///t^ /// sea/t<£ ££///70S A// W0ZZGC? ££££=/* '3 ^ 2 © < do C3 * g« O.-, tf ° o I 2-t* J: % g *VR8 COTTON IN WEEVIL-CONTROL EXPERIMENTS 19 for the weevils to work upon, there is less injury to the bolls, though the fact that from 7 to 9 per cent of the locks on early bolls were damaged shows that the presence of squares is not a complete pro- tection against weevil attack. The greater injury to the later bolls corresponds to the heavier infestation of weevils in the latter part of July and to the fact that there were very few squares on the plants at that time. The increase of about 12 per cent in the injury to bolls in the planting of May 12, as compared with that of May 5, as shown in Table 11, is due to the fact that practically all the bolls produced in the last planting were from flowers appearing after July 8, so that most of the bolls in this planting were immature when the weevil infestation reached its peak during the latter part of July. It is C1 able also that weevil injury caused the shedding of many late that otherwise might have been retained and developed to maturity. YIELDS FROM SUCCESSIVE PLANTINGS AT SAN ANTONIO During the first week in August the first open bolls were found on the plantings of April 19, April 28, and May 5. Although there was little difference in the time when the first open bolls appeared on these three plantings, that of April 19 matured the hulk of its crop sooner than the others. This was in accord with the larger number of early flowers produced by the first planting. The entire crop of all plantings had matured by September 10, but the scarcity of labor delayed picking until September 19. Table 12. — Yield of seed cotton from successive plantings at San Antonio Date planted, 1923 Block Row Number of plants Yield of seed cotton (pounds) Section A Section B Per row Section A Section B Per row Apr. 19 1 8. 25 10 00 18. 25 2 6.75 8.00 14. 75 3 6.25 8. 12 14. 37 4 5.44 7. 25 12.69 f 5 198 194 392 5. 87 7. 87 13. 74 J 6 182 193 375 6.00 7. 75 13. 75 7 205 198 403 4. 75 7. 12 11.87 l 8 189 197 386 4. 82 7. 21 12.03 Block total. 774 782 1,556 21.44 29. 95 51.39 9 194 185 379 5.00 6. 56 11.56 10 190 197 387 5. 00 7.50 12. 50 Apr. 28 1 137 163 300 2. 06 4. 75 6. 81 f 2 126 135 261 2. 19 5.25 7. 44 3 138 147 285 3. 12 6. 25 9. 37 4 162 172 334 3. 58 5. 94 9.52 1 5 131 154 285 3.31 5. 64 8. 95 Block total 557 608 1, 165 12.20 23. 09 35. 28 6 149 172 321 4.00 5. 62 9. 62 May 5 1 170 187 357 3. 25 5.01) 8. 25 { 2 168 177 345 3.70 5. 79 9. 49 8 185 186 371 3.44 4.69 8. 13 4 18K 193 381 4. 52 5. 12 9. 64 l r, 195 181 376 3. 87 4. 18 8. 05 Block total 736 737 1,473 15.53 | 19.78 35.31 0 iW 189 359 4. 00 4. 12 8. 12 20 BULLETIN 1320, U. S. DEPARTMENT OE AGRICULTURE Table 12.— Yield of seed cotton from successive plantings at San Antonio — Con. Date planted, 1923 Block Row Number of plaDts Yield of seed cotton (pounds) Section A Section B Per row Section A Section B Per row May 12 1 220 164 384 2.69 3. 25 5. 94 2 63 80 143 1.00 1. 12 2. 12 f 3 83 105 188 1.00 1.62 2. 62 ] 4 146 135 281 2. 12 2.00 4. 12 1 5 127 149 276 2. 37 1. 18 3.55 l 6 72 131 203 .40 1.37 1. 77 Block total 428 520 948 5.89 6. 17 12. 06 7 107 89 196 .75 .50 1.25 8 99 133 232 1.25 2.37 3.62 1 186 198 384 6.50 9. 06 15. 56 2 193 169 362 5.75 7.25 13.00 f 3 186 184 370 5.00 6. 12 11. 12 4 198 196 394 5.00 5. 75 10. 75 1 5 198 196 394 5. 50 6. 94 12.44 l 6 193 195 388 4.94 7. 37 12.31 Block total 775 771 1,546 20.44 26. 18 46. 62 7 190 194 384 5. 67 6.73 12.40 8 185 207 392 6. 25 7. 37 13. 62 Apr. 28 1 170 174 344 3. 75 4.69 8.44 f 2 182 183 365 3. 06 5.62 8.68 ] 3 160 181 341 3. 50 5.00 8. 50 1 4 162 184 346 4. 05 6. 36 10. 41 l 5 176 144 320 4.00 3.62 7. 62 Block total— — 680 692 1,372 14. 61 20.60 35. 21 6 145 149 294 3.50 4.50 8.00 May 5.-. — 1 166 180 346 4.00 4.37 8. 37 f 2 177 170 347 5.25 4. 37 9.62 J 3 174 174 348 4. 25 3. 62 7.87 1 4 181 173 354 4. 60 5.28 9.88 l 5 178 179 357 4.94 4.25 9. 19 Block total 710 696 1,406 19.04 17. 52 36. 56 6 186 168 354 5. 06 3.87 8.93 May 12 1 173 135 308 3.25 2. 37 5.62 2 112 126 238 1.37 1. 12 2.49 f 3 85 113 198 .50 1.62 2. 12 1 4 114 142 256 2.25 .87 3. 12 1 5 112 162 274 2. 87 2. 06 4. 93 l 6 104 93 197 2.50 1.25 3.75 Block total.- 415 510 925 8.12 5. 80 13. 92 7 100 192 292 2.25 1.75 4. 00 8 178 132 310 2. 44 1.50 3.94 1 7. 19 6. 81 14.00 2 6.00 5.50 11.50 f 3 6.00 6.44 12. 44 ' 4 5. 19 5.00 10. 19 1 5 6. 25 4. 75 11.00 l 6 5. 56 4.62 10. 18 23.00 20.81 43. 81 7 5.75 4.25 10.00 8 5. 62 4. 87 10.49 9 7. 62 4. 75 12.37 10 5. 75 4.00 9.75 COTTON IN WEEVIL-CONTROL EXPERIMENTS 21 The rows in this test were divided into two 100-foot sections by a line stretched across the field, and the weight of seed cotton was recorded from each section of each row. Section A includes the western half of the field and section B the eastern half, where the plants were notably larger. The weight of seed cotton and the number of plants are stated in Table 12. These data are also shown graphically in Figure 7. In comparing the yields of the four plant- ings the weight of seed cotton from only the four inside rows of each block is used. The yields of the outside rows vary to a considerable extent, being affected by adjoining blocks of earlier or later planted cotton. The highest yield of seed cotton was obtained from the April 19 planting, the yields from three 4-row blocks having been 51.39, 46.62, and 43.81 pounds, respectively. The yields from the April 28 and May 5 plantings were about equal. Two blocks of the April 28 planting yielded 35.28 and 35.21 pounds, while the May 5 planting produced 35.31 and 36.56 pounds. Very low yields were recorded from the May 12 planting, the two blocks having produced only 12.06 and 13.92 pounds of cotton, respectively. Consistent differences in the yield of the east and west sections of rows occurred in all plantings. The lower yields from the west sec- tion of rows were due to the presence of Johnson grass, which retarded the growth of plants in this part of the field. In the May 12 planting no thinning was done in parts of rows where a stand was obtained, so that the plant counts are not com- {) arable with those obtained from the other plantings. Though the ow yields of the May 12 planting may be ascribed very largely to the poor stand of plants, the larger proportion of bolls injured by the weevils was also a factor. The higher yields obtained from the cotton planted on April 19 may also be ascribed, at least in part, to the smaller percentage of injured bolls, showing that the weevil conditions were not as severe during the early period of boll development. Nearly the same numbers of flowers and shed bolls were recorded from the April 19, April 28, and May 5 plantings, and the totals of the yields from the four 50-foot sections of rows of each planting from which these records were obtained were nearly the same, having been 12.68 pounds from the April 19 planting, 12.15 pounds from the planting of April 28, and 11.67 pounds from that of May 5. Never- theless, the total block yields were considerably higher on the. April 19 planting. It seems possible that the higher yields obtained from the 50-foot sections of rows in the April 28 and May 5 plantings, as compared with the total block yields, may have been due to the fact that all of the weevil-infested squares were picked up under the 50-foot sec- tions during July for record purposes. The additional protection that may have been given to the 50-foot sections by collecting the weevil-shed squares would not render the yields less significant. On the contrary, the yields that were secured from these sections of rows may be more indicative of the results that might have been secured if the later plantings had been apart from the earliest planting and had not been subjected to the weevil infesta- tion from the early cotton. 22 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE PERCENTAGE OF 5-LOCK BOLLS ON EARLY AND LATE PLANTINGS AT SAN ANTONIO A feature that was found to differ in the several plantings was the proportion of bolls with five locks. A record was obtained of the number of locks in bolls picked from four 50-foot sections of rows in the April 19, April 28, and May 5 plantings. Data from each planting are presented in Table 13, showing the number of 4-lock and 5-lock bolls picked on August 18 and September 10 from two 50-foot rows on the east side of the field and two 50-foot rows on the west side. As stated previously, the presence of Johnson grass on the west side of the field retarded the development of the plants, and lower yields were obtained than on the east side. The data presented in Table 13 show that a consistently higher percentage of 5-lock bolls were produced on the large plants in the east end of the rows. In the April 19 planting the large plants pro- duced 41.2 per cent of 5-lock bolls, as compared with only 23.8 per cent on the small plants. The large and small plants in the April 28 planting produced, respectively, 49.5 and 23.8 per cent of 5-lock bolls and in the May 5 planting 50.3 and 27.6 per cent. There was little difference in the percentage of 5-lock bolls produced on the small plants in the first three plantings, but the percentage of these bolls Eroduced on the large plants in the May 5 planting was 9 per cent igher than in the April 19 planting. Table 13. — Number of bolls and 'percentage of 5-locJc bolls picked on August 18 and September 10 from large plants on the east end of rows and from small plants on the west end of rows in the successive adjacent plantings of cotton at San Antonio Date planted, 1923 Size Date picked Number of bolls Per- centage of 5-lock bolls 5-lock 4-lock 3-lock Total /Aug. 18 \Sept. 10 70 37 162 179 1 1 233 217 30.0 17.1 107 341 2 450 23.8 /Aug. 18 \Sept. 10 146 105 176 181 0 1 322 287 45.3 36.6 251 357 1 609 41.2 /Aug. 18 \Sept. 10 42 51 87 207 1 3 130 261 32.3 19.5 93 294 4 391 23.8 /Aug. 18 \Sept. 10 146 172 99 225 0 0 245 397 59.6 43.3 318 324 0 642 49.5 /Aug. 18 \Sept. 10 36 87 95 225 0 3 131 315 27.5 27.6 Total 123 320 3 446 27.6 /Aug. 18 \Sept. 10 120 178 70 224 0 1 190 403 63.2 44.2 298 294 1 593 50.3 COTTON IN WEEVIL-CONTROL EXPERIMENTS 23 A SEPARATE LATE PLANTING AT SAN ANTONIO A separate late planting of cotton was also made at San Antonio on May 12, the same date as the last of the consecutive adjacent plantings and from the same stock of seed. The surface soil had dried, but by using broad sweeps in front of the planter drill the seed was dropped in moist earth. Germination was rapid, and a fairly good stand of plants was obtained, although most of the rows had a few skips due to imperfect germination. Most of the seed in these skips germinated following a rain on May 30, but the seedlings were weak and most of them died or remained stunted. The location of this planting with relation to other fields of cotton on the experiment farm is shown in Figure 1. CLOSE SPACING IN LATE PLANTINGS % The tendency of late-planted cotton to produce a large “rank” type of stalk under certain conditions renders it desirable to leave the plants closer together in the rows, in order to suppress excessive vegetative growth. Plants which grow large require a longer season to mature a crop, and when the season is shortened by late planting overgrown plants are an added disadvantage. The principle of controlling the vegetative growth of plants by spacing the plants closer in the rows has been tested under a wide range of conditions. Equal or greater yields have usually been obtained from close-spaced plants when tested in direct comparison with wide-spaced plants. A greater degree of earliness is usually obtained by close spacing. In open stands with the plants averag- ing from 2 to 4 inches apart in the row, larger yields have been obtained without thinning. In order to test the effect of close spacing when cotton is planted late, this experiment was planned as a comparison of plants chopped to two plants in a hill with plants left unthinned in the rows. The test consisted of three 4-row blocks of each spacing, the unthinned blocks alternating with the blocks of thinned plants. The outside blocks were protected by guard rows. The plant spacing used in comparison with the unthinned blocks was the same as that used in the time-of-planting test, two plants being left in hills with the hills 12 inches apart. Thinning was done in these rows on June 15, when the plants averaged about 8 inches in height and had from six to eight nodes. OVERWINTERED WEEVILS AVOIDED Although the late-planted cotton was examined at frequent inter- vals for indications of weevil infestation, no trace of weevil injury was found during June. The fact that infestation from overwintered weevils was avoided in the separate late planting probably was due to hot dry weather during the first part of June. During this period the plants were small and had not yet formed squares, so that if weevils came in they had little protection against the high temperatures and did not survive to attack the squares when they had reached sufficient size to enable the weevils to begin breeding. Thus, it appeared that the planting of May 12 had been sufficiently late to avoid infesta- tion from overwintered weevils under the conditions encountered at San Antonio. 24 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE LATER WEEVIL INJURY The first indication of weevil infestation in the late-planted cotton was noted on July 8, after the plants had begun to produce flowers. It is probable that this infestation resulted from migratory or stray weevils from the nearest field of early-planted cotton. A field of cotton planted on April 12 was located about 200 feet southwest of the late-planting test. This early-planted cotton was heavily infested with weevils, and migration may have occurred from this field during the first part of July. In order to show the progress of infestation in the late-planted cotton, a record was obtained of the number of weevil-punctured squares which were shed from the plants in two 50-foot sections of rows. The number of squares shed each day are graphically pre- sented yi Figure 8. The first squares shed as a result of weevil injury were found on July 11. The shedding was very slight until July 25, when a rapid increase occurred. The sudden increase at this time Fig. 8. — Number of weevil-punctured squares shed from a 100-foot section of row in cotton planted on April 22 and on May 12 at San Antonio, Tex. The upper line represents cotton planted on April 22; first squares on May 27. The lower line represents cotton planted on May 12; first squares on June 10. No treatment for the control of weevils was given in either field was probably due to an increased infestation by weevils migrating from near-by fields of early-planted cotton. A similar record of weevil-punctured squares was obtained from a field of cotton planted on April 2, located about 600 feet north of the late-planted cotton. These data are included in Figure 8 for purposes of comparison with data regarding the late-planted cotton. The first squares on the cotton planted April 22 appeared during the last week in May. Weevils were found on plants in this field by June 1, and a 12 per cent infestation was recorded on June 6. As squares were available early in June, it was possible for the weevils to deposit eggs and insure the appearance of a new generation of weevils by the latter part of the month. Thus, the May 12 planting did not get weevils for more than a month later than the April 22 planting, and the breeding of an early generation of weevils during June was entirely avoided in the late-planted cotton. Even if favorable conditions for natural control of weevils did not occur, effective control might be obtained in late-planted cotton by applying poison just before squares are produced. This method of control probably would be limited to late-planted cotton which did COTTON IN WEEVIL-CONTROL EXPERIMENTS 25 not produce squares until the emergence of weevils from hibernation had been completed. Poison applied before the appearance of squares in early-planted cotton would be less effective, as late emerg- ing weevils would reinfest the field. DEVELOPMENT OF PLANTS IN THINNED AND UNTHINNED ROWS In order to compare the development of the plants in thinned and in unthinned cotton, records of plant height and number of nodes on the main stalk, the number of squares produced, and the number of internodes on the fruiting branches were obtained on 10 repre- sentative plants of each spacing. These records were started on June 27 and were taken at biweekly intervals until August 9. The average number of internodes on the mam stalk and the height of plants on June 27 were practically the same on plants which had been thinned to two in a hill on June 15 and on plants left un thinned. Those in both spacings averaged 11.1 nodes on this date, while the height averaged 28.8 centimeters for . un thinned and 28.9 centi- meters for the thinned plants. The number of internodes on plants of both spacings remained practically equal throughout the period during which data were obtained. The final records, obtained on August 9, showed that the unthinned plants averaged 17.2 internodes, while the thinned plants averaged 18 internodes on the main stalk. The thinned plants had grown to a slightly greater size, however, their height averaging 67.3 centimeters, as compared with 63.2 centimeters on those un- thinned. It is apparent that different plant spacings had a negli- gible effect on the development of internodes of the main stalk and on the height of the plants. The dry weather at San Antonio during this season retarded the development of all plants, and it is possible that greater differences in the size of plants would have resulted if more moisture had been available. Although the development of the main stalk was the same on thinned and unthinned plants in this test, the thinned plants had a larger number of internodes formed on the fruiting branches, as shown by records obtained on August 9, when the growth of all plants had practically stopped. These data, showing the compara- tive nodal development of groups of fruiting branches on thinned and un thinned plants, are presented in Table 14. The thinned plants had a consistently larger number of internodes on each group of fruiting branches than the unthinned plants. Table 14. — Number of squares on plants and average number of internodes on fruiting branches of thinned and unthinned plants in a late-planting test of cotton at San Antonio in 1923 Plant spacing Total number of squares on plants at 2-week periods Average number of internodes on fruiting branches on August 9 June 27 July 12 July 25 August 9 Branches 1 to 4 Branches 5 to 8 Branches 9 to 12 10.6 18.7 24. 1 25.4 1.8 2.4 1.7 Thinned (two plants in a hill).. 10. 1 22.9 30.6 34.0 2.3 3.2 2.3 26 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE The average number of squares formed on 10 plants of each spac- ing is shown in Table 14. The unthinned plants averaged 10.6 squares each on June 27, as compared with 10. 1 squares on thinned plants. On August 9, when the final record of all squares formed was obtained, the thinned plants averaged 34 and the unthinned plants 25.4 squares per plant. The greater fruiting capacity of individual plants when two plants are left in hills does not represent the relative fruitfulness of thinned and unthinned cotton. When equal areas of each spacing are com- pared, the greater number of plants in unthinned rows, usually more than offsets the difference in fruiting capacity of individual plants. This is illustrated in Plate III by the number of bolls set on equal sections of rows of thinned and unthinned plants. PRODUCTION OF FLOWERS ON THINNED AND ON UNTHINNED ROWS Daily records of the number of flowers were obtained from July 2 to July 7 and at 2-day intervals thereafter until August 6. These records were obtained from a 50-foot section of rows of thinned and of unthinned plants, each section being representative of similarly spaced plants throughout the test. There were 144 plants in the section of unthinned cotton and 98 plants in the thinned section. The data of flower production are given in Table 15. Table 15. — Flowers counted in 50-foot sections of rows of cotton unthinned com- pared with those on plants thinned to two in a hill with the hills 12 inches apart at San Antonio Date, 1923 Un- thinned Thinned Date, 1923 Un- thinned Thinned Date, 1923 Un- thinned Thinned 30 10 July 13 64 46 July 29__ ... 30 23 July 3 28 8 July 15 69 41 July 31 19 21 22 14 July 17 64 58 Aug. 2 4 17 31 11 July 19 54 48 Aug. 4 3 14 52 17 July 21 46 41 Aug. 6 1 11 July 7 33 22 July 23 41 35 July 9 62 30 43 41 782 566 July 11 50 17 July 27 36 41 Although individual plants of thinned cotton have a greater fruit- ing capacity than unthinned plants, comparisons of equal areas of each spacing show that a larger number of flowers were produced by the unthinned plants. A total of 782 flowers was counted on the 50-foot section of unthinned cotton, as compared with 566 flowers on an equal section of thinned plants. The difference of 216 more flowers on the unthinned cotton represents an increase of about 38 per cent in favor of the unthinned plants in this test. A greater degree of earliness of the unthinned plants is indicated by the number of flowers produced during the first part of the flower- ing period. A total of 372 flowers was recorded on the un thinned plants during the period from July 2 to July 13, as compared with 175 flowers on the thinned plants during the same period. WEEVIL DAMAGE TO BOLLS A record of the extent of weevil injury to bolls was obtained from two 50-foot sections of rows in the late-planted cotton. On August B u ! . 1320, U. S. Dept, of Agriculture Plate 1 1 1 Plants in Separate Late Plantings of Cotton at San Antonio, with the Vegetative Branches Suppressed by Close Spacing A, Two plants in a hill with the hills 12 inches apart; B, unthinned cotton Bui. 1320, U. S. Dept, of Agriculture Plate IV Comparative Size and Fruiting of Cotton Plants in Successive Plantings at Charleston, S. C. A, Planted April 5; B, planted April 25. (Photographed August 14) COTTON IN WEEVIL-CONTROL EXPERIMENTS 27 20 all open bolls were picked, and the number of good and weevil- damaged locks was recorded. Similar data were obtained for bolls which opened between August 20 and September 1 and between September 1 and September 18. A total of 211 bolls was picked from the two sections of rows on August 20. Of these bolls only 25 locks showed signs of weevil dam- age, representing 2.6 per cent of the total number of locks. On September 1, 613 bolls were picked, and 17.9 per cent of the locks were found to be damaged. Of 259 bolls picked on September 18, 34 per cent had damaged locks. A total of 18.8 per cent of the locks was damaged on the 1,083 bolls picked during the season. This weevil injury to bolls was much less than occurred in the other experiment on the cotton planted on the same date but between the earlier plantings. As shown in Table 11, the May 12 planting in the comparison of successive adjacent plantings had 33.9 per cent of all the locks damaged by weevils instead of 18 per cent in the separate late planting. This shows that even a slight isolation of the late plants had a notable effect upon weevil infestation and the resultant injury to the crop. YIELDS FROM THINNED AND FROM UNTHINNED ROWS The late-planted cotton was picked on September 18, at which time all bolls had opened. The field was divided into two equal sec- tions by drawing lines across it at right angles to the rows, and the weight of seed cotton from each section of each row was recorded separately. The length of rows in each section was 100 feet. The row yields from this test are presented graphically in Figure 7 in comparison with the row yields obtained from the successive plantings. The yields and number of plants per row in the late- planting test are given in Table 16. The row yields of seed cotton indicate that soil conditions were very uniform throughout the field. Most of the difference in row yields resulted from imperfect stands, some of the rows having short sec- tions with no plants or with a very irregular stand. This irregu- larity in stand interfered with an accurate comparison of the two systems of plant spacing, as some of the unthinned rows had fewer plants than some rows which had been thinned to two plants in a hill. The poorest stands occurred consistently on the outside rows of each block. As a 2-row planter was used it is probable that these thin stands were due to faulty operation of one side of the planter. In view of the better stands on the two inside rows of each block, a more accurate comparison of yields may be obtained from these rows. The total yield of seed cotton from the inside rows of the three blocks of un thinned cotton was 67.07 pounds, as compared with a yield of 61.97 pounds from the inside rows of the three thinned blocks. From these weights the mean yield of one 200-foot row of un thinned cotton was found to be 11.18 ±0.31 pounds, while the mean yield of an equal length of row of thinned cotton was 10.33 ±0.45 pounds. The difference in average yield of seed cotton between the un thinned rows and the rows which were thinned to two plants in a hill with hills 12 inches apart is less than twice the probable error, indicating that there was no significant difference in yield between the thinned 28 BULLETIN 1320, U. S. DEPABTMENT OF AGRICULTURE and the unthinned cotton. (PI. III.) The cost of production was somewhat lower with the unthinned cotton, however, as the expense of chopping was eliminated. Table 16. — Yield oj seed cotton in a late-planting test at San Antonio in 1928 Unthinned. Do Plant spacing Block Guard— 1 Block total Two plants in a hill (hills 12 inches apart) U Block total Unthinned - 3. Block total Two plants in a hill (hills 12 inches 1, apart) J Block total Unthinned. 5. Block total Two plants in a hill apart) Block total Unthinned (hills 12 inches L Guard— Number of plants Yield of seed cotton (pounds) Section Section Per Section Section Per A B row A B row 14. 00 13. 52 276 282 5.58 6.00 4.81 10.81 320 347 667 6.04 5.62 11. 66 325 321 646 5.25 5.87 11. 12 246 177 423 5.50 4.62 10.12 1, 167 1, 127 2,294 22.79 20.92 43. 71 157 151 308 5.75 5.00 10.75 196 189 385 5.75 4.69 10.44 181 189 370 6.00 5.25 11.25 149 125 274 4.25 4.50 8.75 683 654 1,337 21.75 19.44 41. 19 179 191 370 5.00 4.50 9.50 265 307 572 6. 12 4.69 10. 81 294 281 575 6.49 4.87 11.36 218 216 434 5.25 4.69 9.94 956 995 1,951 22.86 18.75 41. 61 143 158 301 5.50 5.50 11.00 194 195 389 5.20 5. 15 10.35 196 164 360 5.50 5. 19 10.69 155 157 312 4.62 5.00 9.62 688 674 1, 362 20.82 20.84 41.66 196 169 365 4.50 4.62 9.12 269 256 525 5.62 5.00 10.62 260 271 531 6. 19 5.31 11. 50 196 202 398 5.62 5.12 10.74 921 898 1,819 21.93 20.05 41.98 174 140 314 5. 62 5.00 10.62 190 191 381 5.00 5.00 10.00 177 167 344 5. 12 4.12 9.24 100 56 156 5.00 3. 12 8.12 641 554 1, 195 20.74 17.24 37.98 11.50 12.00 11.50 YIELDS FROM COTTON EXPERIMENTS AT SAN ANTONIO The yields of seed cotton obtained from the separate field of late- planted cotton were greater than those from cultural and variety tests conducted at San Antonio during the same season. The yields from the late-planted cotton were exceeded only by the April 19 planting in the successive plantings. The average row yield from the late-planted test, the successive plantings, two cultural tests, and the Lone Star blocks in the variety test are given in Table 17. The same variety of seed was used in all these plantings. COTTON IN WEEVIL-CON TEOL EXPERIMENTS 29 Table 17. — Average row yields of cotton in all experiments conducted at San Antonio Kind of experiment Date planted, 1923 Treatment for weevils Plant spacing Average yield from a 200-foot row (pounds) Separate late planting Successive adjacent plantings. Cultural test No. 1, field C3._ Cultural test No. 2, field C3._ Lone Star check in variety test D3. May 12 (Apr. 19 •J A pr. 28 May 5 (.May 12 Apr. 23 -..do Apr. 22 None. Square stripped, poi- soned. f Unthinned \2 plants in a hill do 11. 18 10. 33 11. 82 8.81 8. 98 3. 25 8.28 6. 78 5.8 6.5 6.5 do None Very poor stand fUnthinned.. .. \1 plant to 12 inches . . fl plant to 10 inches \2 plants to 10 inches.-. SOIL, CLIMATIC, AND WEEVIL CONDITIONS AT CHARLESTON, S. C. A comparison of successive plantings of cotton was made in South Carolina on the farm of F. P. Seabrook on James Island, about 10 miles southeast of Charleston. The soil where the cotton was planted is light and sandy, well drained, and representative of the lighter type of soil of the Sea Islands. It is technically described by the Bureau of Soils as Norfolk fine sand. This type of soil normally produces a comparatively small plant and is admirably suited for cotton. The winters are mild, and the soil becomes warm early in the spring. Cotton can usually be planted the last week in March without danger of frost injury. The summer temperatures are moderate. Rainfall is abundant and fairly well distributed, although periods of dry weather are often experienced in the spring and early summer, while periods of excessive rainfall are common in late July and August. Records of maximum and minimum temperatures and of the precipitation were obtained at James Island from March 14 to October 15. These records are summarized in Table 18. Table 18. — Average maximum and minimum temperatures and monthly precipi- tation at James Island ( near Charleston) , S. C.,from March to October , 1923 Items of comparison Mar. Apr. May June July Aug. Sept. Oct. Average temperatures (° F.): Maximum i 67.3 i 51.5 2.23 72.2 54.6 1.82 77.0 65.8 6. 27 85.6 87.2 88.9 85.4 2 76. 9 Precipitation (inches) _ 2.54 9. 36 8. 09 3. 13 3. 06 1 March 14 to 31 only. 2 October 1 to 15 only. During the season of 1923 cold dry weather during April was unfavorable to the growth of cotton in the seedling stage. With warmer weather and more abundant moisture in May, growth became more rapid, and conditions were favorable for setting a crop during June and the early part of July. Excessive rainfall in the latter part of July and in August resulted in an abnormally high degree of boll shedding. 30 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE Heavy infestation from overwintered weevils is to be expected under normal conditions. The abundant protection afforded by native vegetation and the mild winters of the Sea Island sections afford excellent conditions for successful hibernation. Weevil emergence begins early, the first weevils usually being noted during April, feeding on the growing tips of the young cotton plants. The bulk of the cotton in this district is planted during the latter part of March, and squares begin to appear by the middle of May. During 1923 many fields in this section were heavily infested with weevils by June 10, while other fields farther away from favorable hibernation places showed very slight infestation. Hot dry weather during June caused high mortality of weevil larvae in those fields where clean cultivation was practiced and the plants were still small, so that the sunlight could reach the shed squares. Although weevil infesta- tion was reduced by the natural control factors in June, the appear- ance of a new generation of weevils early in July caused infestation to increase gradually, so that in the fields where control measures were not applied infestation was practically complete by the middle of July. COMPARISON OF SUCCESSIVE ADJACENT PLANTINGS AT CHARLESTON Four successive plantings, as shown in Figure 9, were made on James Island on April 5, 16, 25, and May 4, each planting being in duplicate. The arrangement of the plantings was similar to those at San Antonio. Plats numbered 3, 4, 6, 7, and 8 con- sisted of six rows, while plats 2, 5, and 9 had seven rows, the additional row being con- sidered as a guard, on account of being adjacent to a much earlier planting. Guard blocks 1 and 10 were also planted on each side of the field on April 5. The rows were 310 feet long and spaced 5 feet apart. The experiment was planted with seed of a very uniform strain of the Meade variety of Upland long-staple cotton. This variety has a staple length of 1% inches and is adapted for use on the South Carolina Sea Islands. The seed was planted by hand in hills 12 inches apart. The cold dry weather of April delayed germination and caused a slow growth of the early cotton during the seedling stage. With higher air temperatures and adequate moisture in May, growth became more rapid. The plants in each planting were thinned to two to the hill 12 inches apart when they reached a height Fig. 9. — Flowering record of successive plantings of cotton on James Island, near Charleston, S. C., July 11 to August 17 COTTON IN WEEVIL-CONTROL EXPERIMENTS 31 of about 6 inches. Guard plats numbered 1 and 10 were thinned to one plant per hill 12 inches apart. DAMAGE BY OVERWINTERED WEEVILS While cotton had been grown on an adjoining field in 1922, early destruction of stalks and unfavorable conditions for hibernation in the vicinity of the field undoubtedly afforded some protection to the 1923 planting. Although careful search was made no evidence of weevil infestation was found until June 14, when punctured squares were seen at three places in the field. Records made on June 20 showed that about 1 per cent of the squares had been damaged by weevils. STRIPPING OF FLORAL BUDS The entire field was stripped of squares and poison applied on June 20. The mode of procedure in the removal of squares was essentially the same as that used at San Antonio, described on pages 6 and 7. The laborers were instructed to remove all squares whose bracts were one-half inch or more long, a size that is attained in about 10 days after the young buds are large enough to be distin- guished readily. There were many squares of this size or larger in the April 5 and April 16 plantings, out only a few in the April 25 and May 4 plantings that were large enough to be removed. Counts were made of the total number of squares removed from representa- tive rows of each plat, and the average number of squares removed per plant was found to be as follows: First planting, 4.5; second 3.8; third, 1.6; fourth, 1.1. During the removal of the squares 78 weevils were captured. Field observations indicated that reinfestation from later emerging weevils would have occurred if the squares had been removed earlier than June 20 during 1923. The planting made on April 25 was at the proper stage for stripping on June 20. Plantings made after May 1 could normally be poisoned successfully without stripping, as the weevil emergence was practically complete before the squares were large enough to harbor larvae. Calcium arsenate in dust form was applied to the plants on June 21 with a hand gun. Heavy rains washed the poison from the plants during the afternoon of that day. As poison should remain on the plants a minimum of 48 hours to be effective, a second application was made on June 22, which remained on the plants until June 27. No further measures were taken to control the weevil during the remainder of the season. LATE-SEASON WEEVIL DAMAGE Examinations were made frequently in this test for indications of weevil infestation following the removal of squares and poisoning. On July 12 two small infested spots were found, but infestation re- mained very slight and on July 27 was only about 3 per cent. There was no appreciable difference in the infestation of the different plantings. Migration of weevils from other fields began early in August, and by August 13 the experimental field was heavily infested. No attempt was made to control migratory weevils, as adverse weather conditions and defoliation of the plants by leafworms pre- vented the setting of bolls on any of the plantings after July 31. 32 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE Weevil damage to mature bolls was a minor factor as compared with damage from anthracnose, boll-rot, and bollworms. RATE OF SEEDLING GROWTH INCREASED BY LATER PLANTING In order to compare the rate of plant development in the four plantings, records of the formation of nodes on the main stalk were obtained from 10 representative plants in each planting. Differences in the growth rate during the seedling stage of these four plantings are shown by the interval between the planting date and the date of appearance of the first square in each planting. As shown by the data in Table 19, the period from the date of planting to the appearance of the first square on the April 5 planting was 51 days, as compared with 44 days for the planting of April 16, 40 days for that of April 25, and 38 days for that of May 4. The latter planting reached the fruiting stage of development in 13 days less time than the planting made on April 5. Table 19. — Rate of nodal growth during the seedling stage of cotton grown in succes- sive plantings at Charleston on four different dates Number of days — Average date of appearance of first fruiting branch Date planted, 1923 After planting1 Per node May 26 51 6. 37 May 30 44 5. 50 40 5. 00 June 11 . . 38 4.75 1 As these records were made from the planting date rather than from the date of germination, a deduction of about five or sis days is necessary in comparing these figures with those that were obtained in Texas. (See Table 3.) The April 5 planting required an average of 6.37 days for the development of each of the first eight nodes, as compared with 4.75 days per node for the May 4 planting, a difference in growth rate of 1.62 days. Higher air temperatures and warmer soil conditions dur- ing the seedling stage of growth are probably the factors causing this increased rate of growth in the later plantings. The cotton plant is very sensitive to low temperatures, and if the young seedlings are exposed to periods of cold weather their growth may be so checked that they may not readily resume normal growth 'when conditions again become favorable. PLANT GROWTH DURING THE FRUITING PERIOD Records of plant growth during the fruiting period indicate that the accelerated development of the internodes of late-planted cotton is practically confined to the seedling stage. The average date of appearance of the first and twelfth fruiting branches on 10 plants of each planting and the average interval between the appearance of successive fruiting branches are presented in Table 20. The number of days required by each planting for the development of the first 11 fruiting branches was as follows: April 5 planting, 31 days; April 16, 27 days; April 25, 29 days; and May 4, 28 days. COTTON IN WEEVIL-CONTROL EXPERIMENTS 33 Table 20. — Rate of nodal growth during the fruiting stage of cotton ■ grown in suc- cessive plantings at Charleston on four different dates Date planted, 1923 Average date of appearance of fruiting branches Average number of days for production of fruiting branches Total fruiting branches on Aug. 11 First branch Twelfth branch For 11 branches Per branch May 26 ■_ June 26 31 2. 82 20.3 Apr. 16 .. . May 30 do_ 27 2. 45 21. 5 July 3 29 2. 63 21. 2 July 9 28 2. 54 20. 5 While the rate of fruiting-branch production remained fairly con- stant under conditions of uninterrupted growth during June and early July, the rate of growth in the older cotton was checked about the middle of July, making it possible for the younger cotton to overcome the lead established earlier in the season by the early-planted cotton. The final measurements on all plantings, made on August 11, showed very little difference in the number of fruiting branches. The April 5 planting averaged 20.3 fruiting branches per plant, as compared with 21.5 for that of April 16, 21.2 for that of April 25, and 20.5 for that of May 4. Each node on a fruiting branch provides for the development of a floral bud or square. Thus, the theoretical fruiting capacity of the plant may be measured by the total number of its fruiting nodes. The average per plant of the total number of squares formed on 10 plants of each planting on June 20, Juty 2, July 13, and August 11 are shown in Table 21. The data in this table were afforded by normal unstripped plants. Table 21. — Total number of squares per plant on given dates on cotton grown in successive plantings at Charleston on four different dates Date planted, 1923 June 20 July 2 July 13 Aug. 11 Apr. 5 22. 5 41. 1 52. 6 66. 9 20. 6 39. 0 53. 3 76. 0 Apr. 25. 13. 6 30. 4 45. 6 77.4 May 4 8.3 26.6 42.2 82.7 On June 20 the average number of squares on the plants in each planting was as follows: April 5, 22.5; April 16, 20.6; April 25, 13.6; and May 4, 8.3, a difference of 14.2 fruiting nodes per plant be- tween the first and fourth plantings. The greater fruiting capacity of the April 5 planting was maintained until July 13, at which time it had been passed by the second planting. Final measurements of fruiting capacity were made on August 11. While the total num- ber of fruiting branches per plant in all plantings was practically the same on this date, the fruiting capacity of the different plantings increased progressively with later plantings. The total number of squares produced on the April 5 planting was 66.9, as compared with 82.7 on the May 4 planting. Partial cessation of the growth of the lower fruiting branches in the older plantings rather than a 34 BULLETIN 1320, U. S. DEPARTMENT OP AGRICULTURE slower rate of development is the cause for the complete reversal of the position of the four plantings in regard to the total number of squares formed on the plants. The later plantings continued the production of new internodes on the lower fruiting branches after the growth of the latter was checked on the older cotton. The comparative size and fruiting of early and late planted cotton are illustrated in Plate IV. EFFECT OF REMOVING FLORAL BUDS Records of the number of flowers produced on stripped and un- stripped plants grown in Florida in 1922 indicated that an increased rate of flowering results from the removal of squares from the plants.5 In an endeavor to analyze the reaction of cotton plants to square stripping, comparisons of plant development and the fruiting capacity of a series of stripped and unstripped plants in each of the successive plantings were made at Charleston, S. C., during 1923. From these records it is possible to show the effect of square pruning on the. height of plants, number of fruiting branches, number of internodes on the fruiting branches, and the total number of squares formed on the plants. Squares were removed from 10 plants in each planting on June 20. Table 22 shows the average number of squares removed from these plants, varying from 10.4 squares per plant in the first to 0.8 in the last planting. Table 22. — Number of squares removed on June 20 from cotton plants grown in successive plantings at Charleston on four different dates Average number of squares Date planted, 1923 On plant Squares removed Number Per cent Apr. 5 __ 24.0 10.4 43.3 Apr. 16 _ 21. 1 8.2 38.9 Apr. 25 10.9 2.5 22.9 May 4___ ... _ 7.6 .8 10.5 The height of the plants was recorded weekly from June 21 until August 11. These data are presented in Table 23, showing the weekly growth in height of the stripped and unstripped plants for the four different plantings. The plants in the first three plantings from which squares had been removed on June 20 made slightly more growth during the period from June 21 to August 11. The increase in the growth of the stripped plants was 2.9 centi- meters in the first planting, 17 centimeters in the second, and 4.7 centimeters in the third. The unstripped plants in the last, or May 4, planting made more growth than the stripped plants, the increase being 11.2 centimeters. While the comparison of growth of stripped 1 Smith, G. D. A preliminary report upon an improved method of controlling the boll weevil. Fla. Agr. Exp. Sta. Bui. 165, p. 18-24, illus. 1922. COTTON IN WEEVIL-CONTROL EXPERIMENTS 35 and unstripped plants in the first three plantings shows that the stripped plants grew larger than those not stripped, the increases are hardly significant. Table 23. — Average height of stripped and of unstripped cotton plants grown in successive plantings at Charleston on four different dates in 1923 Average height on date given (centimeters) Planting date and condition June 4 June 14 June 21 June 29 July 6 July 13 July 21 July 27 Aug. 3 Aug. 11 Growth from June 21 to Aug. 11 April 5 planting: Stripped _ Unstripped Increase in growth of stripped 18.6 27.4 35.3 37.3 46.0 46.2 54.7 51.9 62. 1 62.0 68.6 67.4 70.7 70.2 73.3 71.8 74.3 73.4 39.0 36. 1 2.9 Apr. 16 planting: Stripped Unstripped Increase in growth of stripped 21.8 28.2 35.9 33.6 45.9 41.9 55.3 49.2 63.5 53.0 76. 1 58.4 82.4 61.6 85.2 65.4 85.8 66.5 49.9 32.9 17.0 Apr. 25 planting: Stripped.. Unstripped Increase in growth of stripped 14.4 19.2 25.4 27.7 31. 7 33.2 38.7 41. 1 46.9 48.8 58.0 57.4 63.3 61. 2 67.0 64.4 68.7 66.3 43.3 38.6 4.7 May 4 planting: Stripped Unstripped Increase in growth of unstripped 12.0 15.8 21.3 22.8 28.2 31.5 36.7 40.8 47.2 50.6 58.9 63.9 64. 1 70.8 64.0 75.8 65.4 78. 1 44. 1 55.3 11.2 If acceleration of fruiting results from the removal of squares this would be shown by an increase in the number of fruiting branches or by an increased number of internodes on the branches. Weekly records ol the number of fruiting branches from June 3 to August 1 1 on stripped and on unstripped plants are shown in Table 24. In the stripped plants of the April 5 planting, from which 43.3 per cent of the squares were removed on June 20, there was a total increase of only four-tenths of a fruiting branch per plant for the entire seasdn. The gain in the number of fruiting branches of 2.5 in the stripped plants of the April 16 planting and 0.3 in that of April 25 with a loss of 1.4 in the May 5 planting would indicate that no significant in- crease in the number of fruiting branches was obtained by the removal of squares. The stimulating effect of the removal of 43.3 per cent of the squares was not sufficient to change materially the normal rate of frui ting-branch production." 8 Martin, R. D., W. W. Ballard, and D. M. Simpson. Growth of fruiting parts in cotton plants. In Jour. Agr. Research, v. 25, p. 195-208, illus. 1923. 36 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE Table 24. — Average number of fruiting branches 'per plant on stripped and un- stripped cotton grown in successive plantings at Charleston on four different dates in 1923 Planting date and condition Date when data were recorded Number of fruit- ing branches formed from June 19 to Aug. 11 June 3 June 7 ! June 14 June 19 June 26 July 2 July 9 July 13 July 18 July 24 Aug. 3 Aug. 11 Apr. 5 planting: Stripped 4.6 6.1 8.1 10.2 13.0 15.0 16.7 17.4 19. 1 19.9 21.3 21.7 11.5 9.2 11.5 13.7 15.5 16.3 17.7 18.2 19. 7 20. 3 11.1 Apr. 16 planting: Stripped 5.5 6.9 9.6 12.5 14.4 16.4 17.6 19.5 21.0 22.6 24.2 14.6 Unstripped 9.4 11.6 13.4 15.4 16.2 17.5 18.6 20.8 21.5 12.1 Apr. 25 planting: Stripped 2.6 4.5 6.4 9.0 10.7 12.9 13. 9 15.4 17.3 19.6 20.3 13.9 7. 6 10. 0 11.6 13.6 14. 8 16.8 18.2 19.9 21. 2 13. 6 May 4 planting: Stripped... ... .4 2.6 5.3 8.5 9.9 12.4 13.3 15.4 17.0 19. 0 19.0 13.7 Unstripped ... 5.4 8.5 10.0 12.3 13.2 15.3 17.2 19.3 20.5 15.1 The average number of internodes per fruiting branch was ob- tained for the stripped and unstripped plants in each planting. For convenience, these data have been arranged in groups of five consecu- tive fruiting branches and are presented in Table 25. Table 25. — Average number of internodes per fruiting branch of stripped and of unstripped cotton plants grown in successive plantings at Charleston on four different dates in 1 923 Planting date and condition Number of fruiting branches Average total squares 1 to 5 6 to 10 11 to 15 16 to 20 Apr. 5 planting: Stripped _ _ 5.3 4.3 5.0 4.2 4.1 3.2 1.4 1.4 83.6 66.9 Unstripped ___ ______ ___ _ _ __ Increase on stripped plants Apr. 16 planting: Stripped ___ _ _ _ _ 1.0 .8 .9 0 16.7 5.7 4.2 6.2 4.8 4.8 3.8 3. 1 2. 1 105.9 76.0 Unstripped _____ ___ _ Increase on stripped plants Apr. 25 planting: 1.5 1.4 1.0 1.0 29.9 4.9 4.2 5.4 5. 1 4.1 3.9 1.7 2.0 81.6 77.4 Unstripped __ __ .__ ___ _ _ Increase on stripped plants May 4 planting: Stripped., ________ _ _ _ _ _ . 7 .3 .2 -.3 4.2 5.6 5.4 5.6 5.3 4.2 3.8 2.0 1.8 87.0 82.7 Increase on stripped plants __ ___________ .2 .3 .4 .2 4.3 A consistently larger number of internodes was formed on the fruiting branches of the stripped plants in the April 5 and April 16 plantings. As the total number of fruiting branches was practically the same on the stripped and on the unstripped plants the increase in number of internodes on fruiting branches resulted in a greater fruiting capacity of the stripped plants. The stripped plants in the April 5 planting averaged 83.6 squares per plant, while the unstripped COTTON IN WEEVIL-CONTROL EXPERIMENTS 37 plants averaged only 66.9 squares. An even greater increase occurred in the April 16 planting, the stripped plants averaging 105.9 squares as compared with 76 on the unstripped cotton. Only a slight increase in the number of internodes occurred on fruiting branches of stripped plants in the April 25 and May 5 plant- ings. As few squares were removed from these plants, no such stimu- lation of growth would be expected as occurred on plants from which a large percentage of squares was removed. From the foregoing data on the effect of removal of the squares it would appear that no material increase in height or number of fruiting branches was caused. The consistent increase in the number of internodes per fruiting branch indicates, however, that the growth of the fruiting branches is affected by the removal of the early squares. A more continued or prolonged growth seems to result from their removal rather than a faster rate of development. A somewhat analogous though more extreme result of pruning is found in “boll- weevil cotton ” as described by Cook.7 FLOWERING RECORDS OF EARLY AND LATE PLANTINGS Flower counts were started on July 1 1 and were made twice a week until August 17, a period of 38 days. These counts were made on the four inside rows of two plats of the April 5 planting and on one plat of the April 16, April 25, and May 4 plantings. The length of the rows was 310 feet. The flower records for each plat are given in Table 26 and are graphically shown in Figure 9 (p. 30). Table 26. — Flowering record of cotton pla?its grown in successive plantings at Charleston on four different dates Plat Date planted, 1923 Thinning distance Date of counting (number of flowers) Total of 1 1 counts July 11 July 16 July 19 July 23 July 27 July 30 Aug. 2 Aug. 6 Aug. 9 Aug. 13 Aug. 17 No. 2_. Apr. 5 2 plants at 12 inches. 817 1,206 1,463 1,750 1,348 1,496 1, 064 1, 432 1, 172 449 120 12, 317 No. 3.. Apr. 16 do 616 1. 109 1,200 1,519 1, 400 1,428 1,260 1,593 1,384 714 282 12, 505 No. 4.. Apr. 2.5 do 549 976 1,095 1,459 1, 112 1,427 1,200 1,798 1,789 923 335 12, 663 No. 5.. May 4 do 390 841 875 1,264 1,087 1, 358 1, 158 1,689 1, 765 1,218 384 12, 029 No. 6.. Apr. 5 do 728 1,014 1, 184 1, 499 1, 113 1,216 1, 103 1,311 1, 286 593 190 11, 237 Although the first flowers appeared in all plantings at nearly the same date, the more advanced development of the early-planted cotton resulted in a higher flowering rate for this planting during the first part of the flowering period. On July 11, the date on which the first counts were made, 817 flowers were counted on the first planting, as compared with 616 for the second, 549 for the third, and 390 for the fourth. A higher rate of flowering was maintained by the first planting until July 27, at which time a larger number of flowers was recorded on the second planting. Beginning on August 6, the April 25 and May 4 plantings were flowering more profusely than the two earlier plantings. The sudden decline in the rate of flowering which occurred during the second week in August is attributable to infestation from migrat- ing weevils and the defoliation of the plants by the cotton leaf worm. 7 Cook, O. F. Roll-weevil cotton in Texas. U. S. Dej)t. Agr. Rul. 1153, 20 p., illus. 1923. 38 BULLETIN 1320, U. S. DEPARTMENT OF AGRICULTURE In plats 5 and 6, where the April 5 and May 4 plantings were grown side by side, 7 per cent more flowers were recorded from the May 4 planting. Owing to a high rate of boll shedding during late July and in August, few bolls were matured from flowers produced during that time. Data were obtained from 20 plants of each of the four plantings, showing the number and percentage of bolls set from flowers which opened during weekly periods from July 7 to August 14. Data from these 80 plants are combined and presented in Table 27. During the week from July 7 to 14, 80.9 per cent of the flowers were set as bolls. Only 47.1 per cent of the flowers produced during the following week were set and 15.2 per cent during the third week. This declining rate of boll setting continued during the period from August 7 to 14, when only 3.1 per cent of the flowers set. Thus, the larger numbers of flowers produced by the later plantings during late July and August were of little value in setting a crop under the conditions of this experiment. Table 27. — Number of flowers recorded and number and percentage of bolls set on 80 cotton plants at Charleston during each of five weekly periods in 1923 Item July 7 to 14 July 15 to 22 July 23 to 30 July 31 to Aug. 6 Aug. 7 to 14 157 276 296 205 162 Bolls: Number 127 130 45 14 5 80.9 47. 1 15.2 6.8 3.1 YIELDS FROM SUCCESSIVE PLANTINGS AT CHARLESTON As a result of complete defoliation of the plants by leafworms, the bolls opened rapidly during the latter part of August, and the bulk of the crop was open the first week in September. The first picking was made on September 10 and a small second picking on October 4. The field was divided into equal sections by a line across the center of the field at right angles to the rows. Each section of each row was picked separately and weighed on scales graduated to one-tenth of a pound. The picking results are given in Table 28 and graphically presented in Figure 10. Plant counts for each section of row are included in Table 28. Irregular stands resulted in considerable variation in the number of plants per row. While higher yields would be expected in rows having a perfect stand, it has been found impracticable to make cor- rections in the yields on account of deficient stands. The yields from the four inside rows in each plat are used for com- parisons between different plantings. The yields of outside rows were affected by adjoining plats of earlier or later plantings. The total yields from the first and second plats of each planting are as follows: April 5, 279.5 pounds; April 16, 244.9 pounds; April 25, 235 pounds; May 4, 222.4 pounds. While these figures show an increase for the early-planted cotton, reference to the plat yields in Table 28 indicate that most of the gain was due to better soil on one side of the field. Plat 2, planted April 5, yielded 158.5 pounds, while plat 6, planted on the same date, yielded 121 pounds. Plats 3 and 7, Comparative Fruiting of Early-Planted and Late-Planted Cotton in Successive Plantings at Charleston, S. C. The April 5 planting is at the left and the May 4 planting at the right Bui. 1320, U. S. Dept, of Agriculture Plate V COTTON IN WEEVIL-CONTROL EXPERIMENTS 39 planted April 16, yielded 136. S and 108.1 pounds, respectively. Plats 4 and 8, planted April 25, yielded 130.6 and 104.4 pounds, and Nos. 5 and 9, planted May 4, yielded 111.8 and 110.6 pounds, respectively. The fact that such wide differences in yield occurred on the first and second plats of the first three plantings indicates that the high yields from the first plats were due to more fertile soil in that part of the field. The yields from the first and second plats of the May 4 planting were practically equal, and comparatively small differences in the yield of the different plantings occurred on plats 5, 6, 7, 8, and 9. Plat 5, planted on May 4, yielded only 10 pounds less than plat 5, planted on April 5.s (PI. V.) I- ! s * I" n/L jr^CT/O/V & r ,1 "1, / -P’-'l U -lpJ" 1 XT r r'-'L-ri iruj_ ^•^C77 o/v h Pi r j Ir L L— i- \r Vi ^ t srf>£./6 yye.ys sMrr /wry a? / Air and water currents (i Surfaces fl Odorous substances 8 Contact with chemical sub- stances in Discussion , io Conclusions 13 Literature cited 13 INTRODUCTION Insects are generally attracted to materials for three purposes : (1) To obtain food for themselves or their progeny, (2) to lay their eggs, or (3) to gather material for their nests. In some instances the food of the adult and young is the same, and the eggs ape laid directly on the substance which the adult eats. But there are many insects which show no such relation, in which the adult leads some part, often a considerable part, of its life in an environment very dif- ferent from that of its immature forms. Furthermore, certain adult insects do not feed at all, yet are able, in some manner, to deposit their eggs in locations which favor the ready access of the young larvai to their acccustomed food. Indeed, there is so much precision on the paid of many insects in the selection of a place to deposit eggs that students were early impressed with the idea that something directs the gravid female to, and induces her to oviposit upon, food suitable for her progeny. It is the purpose of this bulletin to discuss the various stimuli which affect the oviposition reaction of insects. Any treatment of the subject at this lime must, however, he considered preliminary. Few attempts have been made to analyze this response, although numerous observations are on record which contribute toward its 22803—25 i 2 BULLETIN 1324, U. S. DEPARTMENT OF AGRICULTURE understanding. Many of these records are found in the extensive life-history literature of entomology under titles which conceal then- presence. For this reason, some important contributions have prob- ably been overlooked and, although a sincere effort has been made to cover the ground, completeness is not claimed. The stimuli which determine when and where an insect will ovi- posit begin to operate far back in her life and may continue to affect her till the eggs are extruded. These influences are of two kinds, the internal and the external , and for convenience they will be taken up below in this order. INTERNAL PHYSIOLOGICAL CONDITIONS AFFECTING OVIPOSITION NUTRITION There is evidence to show that the amount and character of the food of an insect affect the production of eggs. An adequate treat- ment of this subject would necessarily involve a discussion of nutri- tion and would lead beyond the limits of the present problem. It is sufficient to say here that numerous authors, including Kellogg and Bell (44)4 Baumberger (6), Glaser (26), and Kopec (4.7) have in- dicated that subnormal nutrition, whether due to the quantity or quality of the food, may have a decided effect on oviposition. AGE Among the groups of insects which possess mature eggs upon reaching the adult stage, some species, under favorable conditions, lay their eggs soon after emerging, whereas others retain them for a more or less extended period of time. The state of nutrition and weather conditions modify greatly the extent of this period (26, 40) • No particular attempt has been made to assemble the literature on this subject and only two references are given here. Breit (10) states that bombycid and noctuid moths lay eggs soon after mating, while most diurnal Lepidoptera fly around a few days before ovi- 2iositing. Age has no influence upon the oviposition of Drosophila melanogaster Meig., provided sexual maturity has been reached (Adolph, 1). FERTILITY Fertility appears to be a stimulus for oviposition in some species, influencing not only the time of egg laying but also the number of eggs deposited. Normal oviposition of the cotton boll weevil (An- thonomus grandis Boh.) apparently will not take place till fertili- zation has been accomplished, but it usually begins soon after that (41)' Mating accelerates the oviposition of Heliothis ohsoleta Fab. (o2). The fertile potato tuber moth ( Phthor'maea opereulella Zell.), according to Graf (29), oviposits within 24 to 48 hours after emergence and most of the eggs are laid within 4 days. The number varies from 38 to 290 eggs, the average, from 114 to 209 eggs, depend- ing upon the nutrition of the female. Contrary to this, virgin females oviposit in from 1 to 7 days after emergence, the average time being 4.4 days. The number of eggs ranges from 1 to 51, with only 22.6 as an average. Unpublished observations of the writer on 1 Reference is made by number (italic) in parentheses to “ Literature cited,” p. 13. THE OVIPOSITION RESPONSE OE INSECTS O Ephestia kuehniella Zell, indicate that oviposition is considerably delayed and the number of eggs reduced if copulation has not taken place. Guyenot (31) and Adolph ( 1 ) obtained evidence fvom. Droso- phila melanogaster that mating is a stimulus for egg-laying; the former thought it was a mechanical stimulus because the first eggs deposited were frequently unfertilized. Picard (61) has also observed this effect in Phthorimaea and Ilesperophanes griseus F. A recent work by Glaser (26) indicates clearly that association with the male sex stimulates egg production in Musca domestica L. and Stomoxys calcitrans L. Virgin females of the imported pine sawfly (Diprion simile Hartig) apparently wait 2 days before oviposition and although they can reproduce parthenogenetically, if not mated they lay only half as many eggs as fertile females (53). Mating is not a factor in the oviposition of many parasitic Hymenoptera (< 3 4, 61). nor in certain social Hymenoptera.2 INTERNAL PERIODICITIES Adolph cites the work of Back and Pemberton (3) on the melon fly (Bactrocera cumrbitac Coq.) to show that internal periodicities may be responsible for the intermittent deposition of eggs by cer- tain species. Such periodic egg-laying occurs in other insects (9) though few references to it have been found. Bishop, Dove, and Parman (8) mention that the house fly ( Musca domestica) lays eggs at 8-day intervals. EXTERNAL INFLUENCES AFFECTING OVIPOSITION TEMPERATURE Temperature influences the rate of many life processes, among which may be counted the activities connected with oviposition. Within the range of each species there is probably an optimum tem- perature for egg-laying. In the alfalfa weevil (Phytonomus posti- cus Fab.) mean daily oviposition follows in general the curve of mean daily temperature (57) ; a similar relation holds for the cotton boll weevil (Anthonomus grandis) (76). A reduction of 3° or 4° C. has been observed to lengthen the oviposition period of T amicus (Ip s) typographicus L. from 1 to 8 clays (35). A cool night retards the oviposition of Hyper a punctata Fab. and it ceases between 7° and 10° C., according to Hudson and Wood (39). A recent study by Detouches (19) on the wax moth (Galleria mellonella L.) shows how markedly temperature may affect the quantity of eggs laid. At 37° 2 8ome additional instances of fertility as a stimulus for oviposition have come to light since the above was written. According to Baker and Davidson (Jour. Agr. Re- - 1 -arch, vol. G, pp. 351-3G0, 1916), tho female of Erionoma pyrioola Baker and Davidson i ails to deposit the winter egg unless fertilized directly after the last integument has h arising from water are strongly attractive to gravid female mosquitoes if uUj: pijiimn Experimentally, he finds dilute aqueous solutions oil methane, hydro- son sulphide, old yeast infusion, and stale urine to be considerably more attractive than water alone. 6 Thus Dietz and Zotek (U. S. Dept. Agr. Bui. 885, 55 up., 1920) find that the ‘•ggs of I he aleurodid Men rwanttiUB woi/hnni Ashby are normally laid on the undersides of the leave:-. The females are negatively heliotropie at the time of oviposition, for when a leaf upon which a female is ovii>ositing is turned over so that the light falls directly upon It. egg laying invariably ceases. 6 BULLETIN 1324, U. S. DEPARTMENT OF AGRICULTURE eluded that two factors were necessary to induce Macroglossum to oviposit, an optical factor, effective at a distance through yellow and green light and a chemical factor operating near at hand through the specific odor of the larval food plant, Galium. Titschack (77) found that the color of wool stuffs is not a factor in deter- mining egg laying in the webbing clothes moth ( Tineola biselliella Hum.) . AIR AND WATER CURRENTS Aquatic and aerial insects are oriented in their environment by the movement of the medium surrounding them. It seems probable that ovipositing insects are also affected by stream or air movement. Wardle (81) states that wind is antagonistic, to the oviposition of blow-flies. The cyrtid fly Pterodontia favipes Gray deposits its eggs on the leeward side of trees (45) , in which location it may be oriented by air movement. SURFACES In many insects, contact with an appropriate surface seems to be a necessary prerequisite for oviposition. According to Loeb (4-9) , a highly developed stereotropism exists in the segments of the repro- ductive organs of animals, and further there are indications that contact with a solid affects the behavior of living matter through an influence on the rate of certain chemical reactions. Crozier and Moore (16) show that the response of cliplopods to surfaces in contact with the body is essentially like the response of a positively heliotropic animal to light; that is, the animal turns its head toward the side which is in contact with a solid surface. When both sides are stimulated by contact with surfaces of equal extent, the movement of the animal is along a straight path. In the cockroach Periplaneta americanci L., contact with suitable material is necessary to bring about the release of the egg case (32) . According to Folsom (25 p. 319), some species of grasshoppers prefer hard-baked soil for oviposition. The migratory grasshopper (Locusta migrator/a L.) in Russia evidences a choice between dif- ferent kinds of soil. Isolated females insert the ovipositor into the soil a number of times before they deposit their eggs, and often a swarm which has alighted on soil too hard for oviposition will re- sume flight again (80) . Baillon (4) also mentions that grasshop- pers choose between different types of soil for oviposition. The Mormon cricket (Andbrus simplex Hald.) is said to prefer a some- what firm but not very hard soil for this purpose (13). According to McColloch (50) the corn earworm moth (Heliothis obsolete ) deposits more eggs on corn plants which have rough hairy stalk and leaf surfaces than on plants with smooth surfaces. The moths were also induced to lay some eggs on cotton twine. Investigations of Benedict (7) and Titschack (77) on the webbing clothes moth (Tineola biselliell-a) suggest that the tactile stimulus may be the determining factor in the selection of a place for egg laying by this species. Any rough surface was observed by Titschack to call forth oviposition, regardless of the food value of the material for the larvae. The moths with which Benedict experimented laid their eggs on cotton and silk as well as wool, the loose threads being es- pecially preferred. The character of the surface is apparently of THE OVIPOSITION RESPONSE OF INSECTS 7 importance to the potato tuber moth (Pthorimgek operculella) . In France, Picard (58) states that it generally lays its eggs in the cavities which surround the buds on the surface of the tuber, in incisions of the skin, or on the clumps of dried earth which adhere to the surface. It will also oviposit on the foliage of Ver- bascimi and Cynoglossum which is felted and plaited, in preference to that of Linaria, for although the latter is more closely allied to the Solanaceae than Cynoglossum, its leaves have smooth surfaces. In laboratory experiments, the moths often laid a part of their eggs on the muslin sides of the cage, even when potatoes were available, but eggs were placed only exceptionally on the glass walls. Graf (29), who has studied the potato tuber moth in America, likewise reached the conclusion that oviposition was stimulated by rough- ened surfaces. The Angoumois grain moth (Sitotroga cerealella Oliv.) does not require the presence of grain as a stimulant for egg laying, but, in captivity, will readily oviposit between strips of cardboard. Usually all the eggs are deposited in the crevice be- tween the strips (75). Dewitz (20, 22), while pointing out the possible role of odor in the attraction of the gravid female of Cocky Us ambiguella , also states that oviposition on the grapevine bud may be attributed to a contact stimulation. In another paper (21 p. 233 ) he quotes Marchal to the effect that the female of Polychrosis hot ran a is guided during oviposition upon the smooth surface of the grape by the tactile power of the abdomen. Ovipo- sition would not take place on grapes covered experimentally with powder or a sticky mass. The experiments of Adolph (1) on Drosophila mekcnog aster show that the texture of the substance with which the gravid female comes in contact exercises a marked effect upon the quantity of eggs laid. Boiled agar was more potent in this respect than any of the solutions which were used to test the effect of taste, odor, or a combination of taste and odor. The character of the nidus also has a very evident influence upon the oviposition of the house fly (Musca domestical). Under appropriate conditions, pine sawdust is considerably less attractive than timothy chaff or horse manure, and moist absorbent cotton (containing ammonium carbonate only) was oviposited upon only once in 11 ex- periments (03, 64) • Some observations by Picard (60) on the ovi- position of Pimpla instigator F., a hymenopterous parasite of the chrysalis of Pieris brassicae L.. and of certain other Lepidoptera, are interesting in this connection. If an old chrysalis shell or a cylinder of white paper is coated with fresh blood from a chrysalis of Pieris, the parasite will pierce it with its ovipositor. The stimu- lus is olfactory, but according to Picard the actual deposition of liie egg depends upon a tactile stimulus produced by the resistance of the living tissue within the chrysalis. Indeed, a chrysalis shell or a hollow cylinder of paper may be many times perforated by the ovipositor, but never will art egg be laid. The importance of tactile stimuli in the oviposition response of Flubrobracon juglamlh lias recently been shown by Ilase (Die Naturwissenschaft, Jahrg. 11, Heft 3*9, pp. 801 806, 1923). Touch is probably the directing sense in the oviposition behavior of Hubroovtus (67). In a recent publication, Howard (36, p. 36-37) declares that the stimulus for oviposition in certain ehalcidoid parasites of gall-mak- BULLETIN 1324, U. S. DEPARTMENT OF AGRICULTURE irig insects is not the morphological character of the host insect hut of the gall which it inliabits. In some other parasites mentioned by this author, the stimulus seems to he furnished by the silken cocoons or webs of the host insects. ODOROUS SUBSTANCES A number of observations are on record which stress to a greater or less degree the importance of odor as a factor in oviposition. Scudder (77), in discussing the so-called botanical instinct of butter- flies, excludes taste and sight but believes the oviposition behavior is in keeping with the idea that the larval food plant is detected by means of the olfactory sense. Tragardh (79) places great emphasis on chemotropism, and Picard (59) also emphasizes its importance but recognizes that light, temperature, humidity, and other physical factors play a part. Brues (11) states that there is much in the behavior of certain species to suggest that food plants are selected by the female insect on the basis of odor. In addition, Brues recog- nizes “some attribute of the plant, perhaps an odor, but far less pro- nounced to our senses than odor or taste '' as a factor in the attraction of insects to plants. Grevillius (30) thought it probable that the choice of a food plant on the part of the brown -tail moth ( Euproctis chfysorrhoea ) was determined by the olfactory sense. The cotton worm moth (Alabama argillacea Hbn.), which lays its eggs on the leaves of the cotton plant ( Gos-sypiurn sp.), may be attracted by the nectar glands on the leaves (13). In fact, moths were seen alter- nately feeding from these glands and ovipositing. It was found, however, that no preference was shown for the portion near the glands on the involucre. This fact induced Comstock to question whether oviposition was here determined by the pres- ence of the nectar glands. Studies by McColloch (59) on Heli- othis obsoleta show that it deposits 60 per cent of its eggs on the silks when the corn plant is in silk. Artificial silks made of cotton twine soaked in the fresh juice pressed from corn silk received 79 per cent of the eggs laid, while the controls (untreated cotton twine) received 21 per cent. Tims odor appears to be important in this case, but surfaces, according to McColloch, must also be considered. Knoll (56) emphasizes the effect of odor upon M acroglossum stella- tarum when the moth is close to the plant upon which the eggs are to be laid. But green or yellow light is necessary to attract the moth to the plant from a distance. The potato tuber moth is attracted by the odor of certain plants (61), but, as previously mentioned, the character of the surface is also highly important. De/witz (30) thought the vine moth (Cochylis avibiguella ) might be attracted and induced to lay its eggs upon or near the buds of the grapevine by the odor poured from the nectaries. But, in addition, he recognized the possible effect of contact stimulation. Loeb (49, p. 160) states that the blowfly 6 is attracted to and will oviposit on decaying meat but not on fat. It will also deposit eggs on objects smeared over with asafetida. A positive chemotropism is responsible, according to this author, for oviposition. Fabre’s observations on the blow-fly, Ccilli- 6 It is here called “ the common house fly,” but the reference is undoubtedly to one of the Calliphoras (cf. Loeb, IS). THE OVIPOSITION RESPONSE OF INSECTS 9 phora vomitoria , indicate that odor is a much more important factor in oviposition than the physical character of ‘the material on which the eggs are laid (23). A variety of substances, colored paper, oil- skin, tin foil, when placed over a receptacle which contained meat, were oviposited upon provided an opening- was made in the cover. Dead birds wrapped in paper envelopes were visited by blow-flies, but they did not lay their eggs on the paper or attempt to oviposit in slits in the paper folds. Fabre attributes this behavior to a mater- nal foresight of the fly for an opening through which the progeny may find their way to food. His results, however, do not preclude the possibility that this behavior resulted from differences in odor con- centration. The same explanation might also be offered to interpret his experiments on the larvipositing fly Sarcophaga carnaria L., (op. cit ., pp. 331-340). Wardle (81) recognizes two factors concerned in the oviposition of blow-flies, (1) the nature of the foodstuffs and (2) meteorological conditions. The stimulus for oviposition, whether olfactory or gustatory he was not sure, probably resides in the exud- ing juices of the food substances. Howlett (37) induced an Indian species of Sarcophaga to deposit larvae in a flask which contained a solution of skatole. Subsequent experiments with skatole by Lodge (Proc. Zool. Soc. London, September-December, pp. 4-81-518, 1916), Roubaud and Veillon (68), and the writer (64) -have failed to sub- stantiate the attractiveness which Howlett claimed for this com- pound. He also obtained eggs of Stomoxys calcitrans upon cotton wool soaked in valeric acid, but an attempt to duplicate the latter result in America failed (65). In the case of the house fly (Musca domesticct) , although the odor of ammonia from ammonium carbonate will, under suitable conditions, induce the female to oviposit (63, 64, 66), the immediate environ- ment from which the ammonia arises also shares in determining whether egg laying will occur. If we place several pieces of solid ammonium carbonate with a little water in a glass dish, female house flies are attracted by the odor, but will not oviposit in or near the dish. A very slight response is obtained with moist cotton and ammonium carbonate which is increased when butyric or valeric acid is added. Pine sawdust is better than cotton but inferior to timothy chaff or acidulated horse manure. Wheat bran is a favorable nidus in the presence of ammonium carbonate, but eggs have not been found in fresh, moist bran which does not volatilize ammonia. It has been shown conclusively that carbon dioxide, a decomposition product of ammonium carbonate, is not in itself attractive to the gravid female house fly but, together with other factors, may exert an influence upon oviposition which has not been detected (17, 18, 66, 68). Adolph (1) found that odor is a slight stimulus to egg lay- ing in Drosophila melanog aster, being most marked when flies could gain contact with the odorous solution. Texture, however, was more effective than odor, and suitable combinations of texture and odor (the flies were prevented from reaching the odorous sub- stance) gave responses nearly equal to those which prevail under natural conditions. Townsend (78), in a study of the tachinid flies, observed that Eupelcteria magnicornis Zett., which deposits living larva* on the foliage of plants, seeks for this purpose only those por- tions over which the host caterpillars have crawled. The parasitic 2280&— 25 2 10 BULLETIN 1324, U. S. DEPARTMENT OF AGRICULTURE larvae are usually placed on stems where a silken thread has been left by a caterpillar, and Townsend suggests that the sense of smell induces the flies to larviposit in such locations. Picard (60) states that the functioning of the ovipositor of Pim'pla instigator is a re- flex determined by an olfactory sensation, but that the tactile sense governs the actual deposition of the egg in the host. The investiga- tions of Hase show that odor is all important in the discovery of the host fly Habrobracon juglanclis but that tactile stimuli are neces- sary to bring about the deposition of the egg (3 Ip.; cf. also Die Natur- wissenschaft, Jahrg. 11, Heft 39, p. 801, 1923). CONTACT WITH CHEMICAL SUBSTANCES In addition to the effects produced by the purely physical char- acter of surfaces there yet remains the possibility that the oviposi- tion behavior may be influenced by direct contact of the insect’s flody with chemical substances. Responses due to the sense of taste and to the general chemical sense proflably belong here. Mclndoo (51, 52) believes that the senses of smell and taste in insects are inseparable. Minnich (5 If, 55, 56), however, has recently described a chemical sense analogous to taste located on the tarsi of two species of Lepi- doptera, Pyrameis atalanta L. and Vanessa antiopa L. Experi- ments on Drosophila, melanog aster (1) indicate that the taste of an aqueous glucose solution is much more effective in evoking oviposi- tion than the odor of a solution which contains a mixture of acetic acid and alcohol, although the latter mixture has a marked food at- traction for this fly (5) . Sharma and Sen (72) , in a study of the ovi- position of mosquitoes, find that weak solutions of sodium chloride, sodium citrate, sodium tartrate, and certain other salts are conducive to egg laying, while the corresponding acids are repellent. Observa - tions of Hancock (33) on the oviposition of the grasshopper Orcheli- mum glaberrimum Burm. reveal the interesting fact that this insect, when searching for a place to lay its eggs, either ignores the plants distasteful to it or subjects them to a brief mouth test (cf. If, p. 13Jf). Although not proving the point, these observations suggest that taste plays a part in the selection of the plant. Brues (11) places taste among the senses which direct gravid female insects to plants. DISCUSSION Insects which spend most of their lives upon substances that offer food for themselves or their offspring probably exhibit the simplest oviposition responses. When the internal physiological conditions are right, simple contact with the stimulating medium appears to be all that is necessary to release the eggs. The behavior of the ovi- positing queen bee suggests that the response is largely determined fly the tactile sense and this may also be true of other colonial in- sects. The webbing clothes moth (Tineola biselliella) , which ovi- posits as readily upon the surfaces of indigestible materials as upon the natural food of its larva, likewise seems to lay its eggs largely in response to tactile stimuli. Contrasted with these simpler cases, the oviposition response of many active free-living species is much more complex. The inten- THE OVIPOSITION RESPONSE OF INSECTS 11 sity and wave length of light, temperature, and humidity, rate of movement of the medium in which she lives, odor, and the physical and chemical character of surfaces aid in bringing the gravid female insect into contact with the specific larval food and induce her to re- lease the eggs. A given set of stimuli is not effective for all species. Thus, for Drosophila melanogaster the stimuli may be roughly clas- sified in the following ascending order of effectiveness: Odor, mois- ture, taste, odor and taste, texture, texture and odor, and a combi- nation of texture, taste, and odor. In comparison with Drosophila, the house fly is more dependent on the odor of the medium; most substances which do not liberate ammonia probably seldom, in na- ture, evoke egg deposition. The response of Macroglossum to green and yellow light presents a reaction at present apparently unique among insects, but which further study may show to be widespread in those species which lay their eggs on green plants. The experimental evidence at hand suggests, then, that a chain of stimuli is, in many species, necessary to induce egg deposition. Adolph (1, p. 338) sets forth this view in the following words : Egg laying in its nature is a complete response (“all or none”) ; that is, partial stimulation can not be measured. A single potent factor in the chain may never lead to the extrusion of eggs. A similar view is gained by Knoll (46 1 p. 349) from his study of Macroglossum, by Picard (60) from observations on Pimpla insti- gator F.. by Hase from studies on Habrobracon juglandis , and the results of the writer’s experiments on the house fly are concordant with it. Loeb (49) seems to favor the idea than an odor stimulus is suffi- cient to produce oviposition in certain free-living insects. He says (p. 160) : The fact that eggs are laid by many insects on material which serves as a nutritive medium for the offspring is a typical instinct. An experi- mental analysis shows again that the underlying mechanism of the instinct is a positive chemotropism of the mother insect for the type of substance serving her as food ; and when the intensity of these volatile substances is very high, that is, when the insect is on the material, the egg-laying mechanism of the fly is automatically set in motion. Thus the common house fly [see footnote, p. 8] will deposit its eggs on decaying meat, but not on fat; but it will also deposit it [them] on objects smeared over with asafetida on which the larvae can not live. * * * It seems that the female insect lays her eggs on material for which she is positively cliemotropic, and this is generally material which she also eats. Fabre’s observations on the blow-fly C alliphora vomitoria empha- size the predominance of odor in this response, and Howlett’s re- sults with Sarcophaga would appear to leave little doubt that odor alone can induce insects to oviposit. It must be said, however, that Howlett’s experiments are given in little detail and might be acci- dental or unusual rather than the customary response of the fly in question. And the observations of Loeb and Fabre do not exclude effects due to the surface with which the flies came in contact. At all events, it is desirable that thoroughgoing evidence be obtained before accepting as fact the proposition that free-living Diptera can be induced to lay eggs solely by means of an odor stimulus. It seems necessary to stress the dependence of chemotropism upon other fac- tors at this point because certain entomologists have rather accepted 12 BULLETIN 1324, U. S. DEPARTMENT OF AGRICULTURE it as the stimulus responsible for the oviposition of insects. From present knowledge, however, it seems doubtful whether a free- living insect can ever be induced to oviposit by means of an odor stimulus alone. The reaction of Drosophila to odor concentration is interesting. It has been shown by Adolph (1, p. 334 , 335) that odor concen- trations are never so low that they fail to call forth positive re- sponses, and even very faint odors have full stimulating value. If this proves true of many insects it will perhaps explain how the faint odors emanating from the green portions of some plants may possess great stimulating value, particularly when the insect is near by. In captivity, some species will oviposit on almost any convenient surface, but others hold strictly to specific substances and refuse to oviposit in their absence. Among Lepidoptera, for example, there are species (Satyrus dry as, Carpocapsa pomenella ) which lay their eggs at random on the walls or floor of the cage, and others which refuse any but a particular food plant for this purpose ( Papilio machaon L.. Pieris hrassicae ., Arginnis selene Schiff., and others, 24, 4% , 69). These results show the difference in oviposition behavior that may occur in the same family of insects. It has been observed, however, that there are occasional errors of judgment on the part of female insects which have specialized food plants; that eggs are, in fact, sometimes placed upon plants which can not nourish the larvae. Knoll (46) observed the habits of Macroglossum stellatarum in captivity, the larva of which is closely restricted to plants of the genius Galium. After retaining the eggs a long time, the female will deposit them on any avail- able green portion of a plant, regardless of its botanical relation- ships. And more recently Schwarz (70) concludes from observa- tions on Catocala extending over a number of years that such mistakes in oviposition are a phenomenon of old age and a sign of physical exhaustion. The question now arises, how has the female insect obtained the ability to respond to these stimuli which lead it almost unerringly to the specific larval food? Is it impelled by a series of tropisms, or by an instinct which is the result of natural selection, or by an acquired instinct now hereditarily fixed? The tropistic view has been advanced by Loeb (48, 49), Tragardh (79), Howlett (37, 38), and others. Brues (11) and Loeb (49, p. 160) have mentioned the possible relation qf natural selection to food selection by the female insect. Bachmetjew (2) believes that the female insect must have an acquaintance, with the taste of the larval food plant which it has inherited from the larva. To use his own words (p. 713). Allein tier Gerueksempfindung bei der Wahl der betreffenden Pflanzeging die Geschmacksempfindung gesckiehtlich voran, denn um zu wissen, wo er seine Eier ablegen soil, musste der Falter zuerst mit dem Geschmack der betreffen- den Pflanze bekannt gewesen sein, resp. dies von der Raupe geerbt kaben. Wheeler (82, p. 71-72) states that oviposition and feeding upon the host blood in the parasitic Hymenoptera are congenitally or hereditarily conditioned reflexes. Little of an exact nature seems to have been done to elucidate this important question. However, the very suggestive experimental investigation of Craighead (14, 15) THE OVIPOSITIOlSr RESPONSE OF INSECTS 13 throws considerable light upon it.7 Craighead finds that nearly all adult cerambvcids display a marked preference for the host wood m which they have fed as larvae, and that certain species which can be induced to feed in a new host show a preference for that host when they become adults. Concern i no; opposition, he says (Ilf. p. 220) : Although the adults show a decided predilection for a favored host in ovipositing and even, in certain species, a preference for the plants in which the larvae have fed for one or two generations, the instinct to oviposit seems to overbalance that of host selection, consequently new hosts are frequently selected — possibly more frequently in nature than is generally realized. If it can be ghown that the food of the larva determines the host preference of the adult, a decided step in advance will have been made. Another step then will be to explain whether the “ memory ” of the food plant which the larva has passed on to the adult is the result of or is influenced by the chemical or physical effects of the food in the growing larva. CONCLUSIONS The following internal factors may condition the opposition responses of insects : The nutritive state as affected by the amount and chemical constitution of the food, age, fertility, and internal stimuli which determine periodic egg-laying. The external influences which may affect the oviposition response are temperature, humidity, light (including color), air currents (and probably in some species water currents), the physical character of surfaces, the chemical constitution of substances which stimulate on contact, and the volatile constituents of substances. The simplest opposition responses are probably shown by insects which spend most of their lives upon substances that serve as food for themselves and their offspring. Most free-living insects, however, require a'* chain of stimuli to provoke egg laying; a single stimulus is insufficient to call forth a normal response. Many species demand a specific chain of stimuli. The odor of a substance may attract gravid female insects, but is probably never in itself sufficient to induce normal opposition. LITERATURE CITED (1) Adoi.ph. Edwabu I*\ 1920. Egg-laying reactions in the pomace fly, Drosophila. In .Tour. Expt. Znol., vol. 31, pp. 327-341. (2) Bachmet.iew, P. 1007. Experl mentelle entomologische Studien. vol. 2, xvi-t-044-f-cviii pp., 01 figs., Sophia. (3) Hack, E. A., and Pemheuton, C. E. 1014. Life history of the melon fly. In .Tour. Agr. Research, vol. 3, pp. 200-274. ( \) P.AirXON, P. Cappk de. 1020. Contribution anatomique et physiologique il l’etude de la repro- duction ciiez les locustiens et les grilloniens, T. La ponte el: lY-elosion ciiez les locustiens. In La Cellule, vol. 31, pp. 1-245. ( 5 1 Harrows, Wim.iam Morton. 1007. The reactions of I lie pomace fly. Droxoph il OFFICE OF INFORMATION DIVISION OF PUBLICATIONS DEPARTMENT BULLETINS Nos. 1326-1350 WITH CONTENTS PREPARED IN THE INDEXING SECTION UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON / CONTENTS Department Bulletin No. 1326. — Effect of Garlic on the Flavor page and Odor of Milk: Object of the experiment 1 Details of experiment 2 Milk tests 2 Check samples 2 Time required for garlic flavor and odor to enter milk 4 Time required after consumption of garlic for the milk to be free from garlic flavor and odor 5 Inhalation of garlic 7 Blood tests 9 Conclusions 9 Department Bulletin No. 1327. — Production of Grape-Hyacinth Bulbs: General characteristics of grape hyacinths 1 Sources of supply 2 Uses of the plants 2 Varieties and their commercial application 2 Detracting characteristics 4 Who should grow the stocks? 4 Soils 4 Planting and cultivation 5 Digging or harvesting 5 Importance of proper drying of bulbs in storage 6 Packing for shipment 7 A possible advantageous method of growing 7 Preparation and sizing of stocks 8 Planting stocks 8 Merchantable stocks 9 Propagation 9 Growing from seed 10 Prime difficulty with grape hyacinths 11 Where grape hyacinths may be grown 12 Present demand 13 Growing under glass 13 Yields 14 Future of grape hyacinths 14 Summary 15 Department Bulletin No. 1328. — The Flight Activities of the Honeybee: Introduction 1 The apparatus 2 Problems on which information may be obtained by a device for counting flights ^ 6 The conditions of the experiment 7 Factors introducing error in the count 8 Factors influencing the flight 10 The average duration of trips 26 A limit to the number of trips and the time spent within the hive 32 The death rate of the colony 33 The behavior of the bees to the instruments 34 Conclusions 35 794 09H— 28 3 4 DEPARTMENT OF AGRICULTURE BULS. 1326-1350 Page Department Bulletin No. 1329. — Bamboos: Their Culture and Uses in the United States: Introduction 1 Botany of the bamboos 4 Cultural types 6 Giant timber bamboo 7 Stake and forage-crop bamboo 8 Dwarf hardy bamboo 9 Hairy sheath edible bamboo 10 Smooth-sheath edible bamboo 11 Variegated Phyllostachys 11 Calcutta fish-rod bamboo 12 Naked-culm clump bamboo 13 Available and prospective types 14 Uses of bamboo 14 Farm-home groves 16 Bamboos and poultry 19 Bamboos and rural school grounds 19 Windbreaks, hedges, and screens ' 20 Bamboos as forage and grazing crops 20 Bamboos for edible purposes 20 Commercial uses 21 Bamboos for timber 23 Bamboos as paper-making material 25 Minor uses 25 Propagation and culture 26 Bamboos as ornamentals 33 Bamboo diseases 37 Bamboo smut 37 Bamboo rust 37 Melanconium culm disease 38 Bamboo insects 39 Bamboo scale 39 Aster olecanium miliaris longum 40 Cottony bamboo scale 40 Leucaspis bambusae 41 Long-tailed mealybug 41 Bamboo plant louse 42 Bamboo mite 42 Prionid root borer 42 Japanese sheath mite 42 Literature cited 45 Department Bulletin No. 1330. — Abbreviations Employed in Experiment Station Record for Titles of Periodicals: Abbreviations for titles i Abbreviations for single words 146 Department Bulletin No. 1331. — The Madonna Lily: History . 1 Varieties 2 Sources of planting stocks 2 Propagation 2 Making a seed crop 7 Soils . 8 Soils that heave should be avoided 8 Planting 9 Cultivation 10 Digging 10 Handling the bulbs 10 Where the lily may be grown 11 Climatic advantages 11 The bulbs after forcing 12 Commercial sizes 12 Packing and shipping 12 Transplanting easy 13 Methods of culture applicable to the Nankeen lily 13 t; CONTENTS 5 Department Bulletin No. 1331. — The Madonna Lily — Continued. page Advantages of home production 13 Enemies 13 Growing under glass 15 The Madonna lily in the garden 15 . Garden associations 16 Present consumption 16 Summary 17 Department Bulletin No. 1332.— Emulsions op Wormseed Oil and op Carbon-Disulfide for Destroying Larvae op the Japanese Beetle in the Roots op Perennial Plants: The plants concerned. 1 Preliminary work 2 Oil of wormseed (American) 3 Wormseed-oil emulsions 4 Toxicity of wormseed-oil emulsions 7 Application to larvae in soil and plants 9 Value of wormseed oil as an insecticide 12 Treatment of peony roots 12 Carbon-disulfide emulsions 13 Toxicity of carbon-disulfide emulsion to larvae 14 Application of carbon-disulfide emulsion to larvae and peonies 15 Commercial use of emulsions 15 Summary and conclusions 16 Literature cited 17 Department Bulletin No. 1333. — Fattening Steers on Velvet Beans: Outline of experimental work 1 The experiment at Collins, Miss 5 The experiments at Beltsville, Md 8 The experiments at McNeill, Miss 17 Conclusions based on experimental work 26 Department Bulletin No. 1334. — Tests op Barley Varieties in America: Scope of the bulletin 1 Barley growing in America 2 Preparation of material 13 Discussion of station yields . 17 Regional adaptation of varieties 154 Well-known or promising varieties 159 Accession data of cereal investigations numbers 175 Botanical comparison of prominent varieties 205 Best varieties of the period from 1917 to 1921, inclusive 207 Distribution of varieties by experiment stations 212 Index 213 Department Bulletin No. 1335. — Commercial Dehydration op Fruits and Vegetables: Preservation by dehydration 1 Dehydration industry 2 Dehydration plant 2 Selection of material 4 Preparation of material 4 Washing 4 Grading for size 4 Peeling 5 Trimming 6 Checking 6 Subdividing 6 Pitting and seeding 7 Traying 8 Conveying the trayed material 8 6 DEPARTMENT OF AGRICULTURE BULS. 1326-1350 Page Department Bulletin No. 1335. — Commercial Dehydration op Fruits and Vegetables — Continued. Pretreatment 8 Blanching or processing 9 Sulphuring 10 Drying 10 Equipment 10 Heat 13 Air 15 Moisture in the air 15 Relation of drying conditions to drying rate and quality of product 16 Recirculation 16 Engineering calculations for designing a drier 21 End point of dehydration 25 Curing 26 Insects attacking dried fruits 27 Preventive measures 27 Remedial measures 28 Packing and storings. 29 Detailed directions for drying 31 Fruits 31 Vegetables 35 Bibliography , 39 Dehydration of fruits and vegetables 39 Control of insect pests 40 Department Bulletin No. 1336. — Biological Studies of the Green Clover Worm: Introduction 1 Systematic history 1 Synonymy ■ 1 Geographical distribution 2 Food plants 2 Description of stages 3 Life history and habits 8 Seasonal history 14 Natural enemies 16 Literature cited 19 Department Bulletin No. 1337. — Work op the Northern Great Plains Field Station in 1923: Introduction 1 Arbori cultural investigations 2 Cooperative shelter-belt demonstrations 2 Experimental tree plantings 6 Horticultural investigations 8 Pomological investigations 8 Oleri cultural investigations 11 Ornamental horticulture 12 Agronomic investigations 13 Rotation and tillage experiments 14 Cooperative cereal and forage-crop investigations 15 Conclusions from agronomic investigations 15 Grazing investigations 16 Department Bulletin No. 1338. — -The Family Living prom the Farm. Data prom 30 Farming Localities in 21 States for the Years 1918 to 1922: Significance of family living from the farm 1 Localities studied 2 The family living from the farm 6 Years of prosperity and depression 11 Cost of living of farm families 12 The farm business 13 Size of farm 10 Size Of family_ 21 CONTENTS 7 Page Department Bulletin No. 1338. — The Family Living from the Farm. Data from 30 Farming Localities in 21 States for the Years 1918 to 1922 — Continued. The farm business — Continued. Farm receipts 23 Farm income 25 Family income 26 Labor income 26 Value of the farmer’s labor. 27 Tenure 27 List of references 28 Department Bulletin No. 1339. — The Effect of Weather Upon the Change in Weight of a Colony of Bees During the Honey Flow: Introduction 1 Method of obtaining data 5 Method of presenting data 8 Comparison of changes in weight of two colonies of bees 11 The spring period 15 The fall period 22 Correlations between external factors and the changes in colony weight :■ 25 The effect of unknown factors on changes in colony weight 38 Theoretically changing weather factors and predicting resulting gains 44 Conclusions 48 Literature cited 50 Department Bulletin No. 1340. — Irrigation Requirements of the Arable Lanes of the Great Basin: Introduction 1 Units and forms of expression 3 The Great Basin 4 General character of the soils of the Great Basin 5 Climatic conditions 6 Water supply 6 Agricultural products 12 The relation of irrigation water to crop production 13 Time of irrigation 19 •Conditions influencing the quantity of water required for irrigation. _ 29 Water requirements as affected by State, community, and corporate regulations 33 Lands to be reclaimed 37 Seasonal net water requirements 38 Use of water on crops in the Great Basin 39 Appendix 41 Department Bulletin No. 1341. — Effects of Continuous Selec- tion for Ear Type in Corn: Object of the investigation 1 Material and methods 2 Establishing the strains 2 Growing seed for comparison 2 Methods of comparison 2 Experimental data 3 Data on number of kernel rows 5 Data on productiveness 6 Discussion 7 Summary 10 Department Bulletin No. 1342. — Effect of Feeding Green Rye and Green Cowpeas on the Flavor and Odor of Milk: Experimental feeding of green rye 2 Procedure 2 Feeding 15 pounds one hour before milking 4 Feeding 30 pounds one hour before milking 4 Feeding 30 pounds immediately after milking 5 8 DEPARTMENT OF AGRICULTURE BULS. 1326-1350 and Green Cowpeas on the Flavor and Odor of Milk — Continued. Experimental feeding of green cowpeas 5 Procedure 5 Feeding 15 pounds one hour before milking 6 Feeding 30 pounds one hour before milking 7 Feeding 30 pounds immediately after milking 7’ Conclusions 8 Department Bulletin No. 1343. — -Improved Oat Varieties for the Corn Belt: Introduction 1 History and methods of oat experiments 2 Albion (Iowa No. 103) 3 History and description 3 Yields of Albion 5 Ratio of grain to straw 9 Rate-of -seeding experiments 9 Richland (Iowa No. 105) 10 History and description 10 Yields of Richland 11 Comparative yields of Richland and Albion 13 Rate-of -seeding experiments 14 Iowar 15 History and description 15 Yields of Iowar 17 Ratio of grain to straw 19 Rate-of-seeding experiments 20 Iogren 20 History and description 20 Yields of Iogren 21 Rate-of-seeding experiments 23 Yields of other varieties at the Iowa station 23 Yields of Albion, Richland, and Iowar oats outside of Iowa 25 Summary 29 Department Bulletin No. 1344. — Effect of Various Factors on the Creaming Ability of Market Milk: Reasons for investigation 1 Methods used 3 Effect of pasteurizing market milk 3 Effect of heating milk to various temperatures • 4 Effect of heating and holding milk for 30 minutes at various tempera- tures at pasteurizing plants _ 5 Effect of pasteurizing fresh milk 7 Effect of pasteurizing old milk 7 Effect of partially cooling pasteurized milk in vat or tank 9 Effect of the temperature to which milk is cooled after pasteurization _ 10 Effect of storage temperature 11 Rapidity of creaming of raw and pasteurized milk 13 Effect of age and recreaming of raw milk 14 Effect of recreaming pasteurized milk 15 Effect of pumping milk - 16 Effect of holding milk and of agitating it at various temperatures before and after pasteurization 18 Effect of clarifying raw milk 21 Effect of filtering milk 21 Influence of various shapes of milk bottles on depth of cream layer- _ 22 Summary and conclusions 23 Department Bulletin No. 1345. — Saltbushes and Their Allies in the United States: Introduction 1 The saltbushes and their allies 2 Results of examination and analysis 3 Goosefoot family 3 Pigweed family 28 Buckwheat family 30 Common names and their scientific equivalents 37 Literature cited 38 CONTENTS 9 Page Department Bulletin No. 1346. — Status of the Pronghorned Antelope, 1922-1924: The pronghorned antelope 1 Former and present abundance of pronghorns 1 Characteristics of the American antelope 4 Chosen habitat 7 Conservation and control 8 Conservation organizations and the antelope 9 Washington conference on the conservation of the pronghorn 11 Establishment of antelope refuges in Nevada 14 Proposed Owyhee Antelope and Sage-hen Refuge, Idaho 15 Restocking experiments, 1924 16 Methods of capturing and transplanting antelope 18 Results of a census of existing antelope 22 Department Bulletin No. 1347. — Foot-Rot Diseases of Wheat in America: Introduction 1 Use of the term “foot rot” 2 Take-all, the foot rot caused by Ophiobolus graminis : 3 The foot rot caused by Helminthosporium sativum 17 An undentified foot rot found in the Pacific Coast States 28 Other foot rots 30 Other fungi capable of causing foot rots 33 Discussion 34 Literature cited 37 Department Bulletin No. 1348. — An Appraisal of Power Used on Farms in the United States: Introduction 1 Sources of information 5 Sources of power used on farms 6 Annual use and cost of power on farms in the United States 7 Number of power units or installations on farms and number of workers engaged in agriculture 9 Primary power available and horsepower-hours utilized annually on farms 10 Effect of the use of power and machinery on production and income. 14 Power and labor requirements of farm operations 17 Power and labor requirements of farm commodities 19 Distribution of farms and farm lands and types and sizes of farms 19 Seasonal distribution of the use of labor and power on farms 20 Factors affecting the efficient utilization of power and labor on farms. 42 Choice of power 45 The future use of power on farms 50 Appendix I. Tables 53 Appendix II. Selected bibliography 73 Department Bulletin No. 1349. — The Brood-Rearing Cycle of the Honeybee: Introduction 1 Method 4 Annual brood-rearing cycle 6 Description of the colonies used in 1921 12 Seasonal characteristics of 1921 13 Brood rearing of a typical colony for two successive seasons 14 General observations on the remaining colonies 16 General discussion of the records for 1921 25 Observations in 1920 27 Migrations of the queen within the hive 30 Compactness of brood nest 32 Time relation of brood rearing to nectar gathered 33 Egg laying... 35 Conclusions 36 Literature cited 37 Tables 38 Graphs 44 10 DEPARTMENT OF AGRICULTURE BULS. 1326-1350 Department Bulletin No. 1350. — Blue-Fox Farming in Alaska: page Introduction 1 What is a blue fox? 3 Brief history of blue-fox farming 4 Fox-growing areas in Alaska 5 Selecting an island or ranch site 7 Climate and shade 7 Location and soil 8 Harbor facilities 9 Food supply 9 Water 9 Island area 10 Ranch organization 10 Structures . 10 Trap-feed houses 12 Breeding stock and equipment 15 Essentials of breeding 16 Pelts 17 Conformation 18 Breeding 18 Time of breeding 19 Mating 19 Essentials of feeding 19 Kinds of feed 20 Methods of preparing and feeding 21 Quantity and frequency 23 Transportation 24 Pelting 25 Primeness 25 Killing 26 Skinning 26 Drying pelts 27 Characteristics of a good pelt 28 Losses from depredations 29 Sanitation and treatment of disease 29 Failures and abandonments 31 Breeders’ associations and ranches 32 White-fox farming in northern Alaska 32 Publications relating to fur animals 31 UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1328 Washington, D. C. ▼ May, 1925 THE FLIGHT ACTIVITIES OF THE HONEYBEE1 By A. E. Ltjndie, formerly Field Assistant, Bee Culture Laboratory, Bureau of Entomology CONTENTS Page. Introduction 1 The apparatus 2 Problems on which, information may be obtained by a device for counting flights 6 The conditions of the experiment 7 Factors introducing error in the count 8 Page. Factors! influencing the flight 10 The average duration of trips 26 A limit to the number of trips and the: time spent within the hive 32 The death rate of the colony 33 The behavior of the bees to the instruments 34 Conclusions 35 INTRODUCTION Although the flight of honeybees to and from the hive has attracted the attention of students of beekeeping from the earliest times, no detailed study has been made, so far as can be determined from the beekeeping literature, of the actual number of flights from a colony of bees or of the variations which occur in these flights with changes in external environmental factors. Obviously it is extremely diffi- cult to obtain even a small number of accurate records by counts made at the entrance of the hive. Therefore, to obtain adequate scientific data for a thorough study of the problems pertaining to the flight of bees it is essential to have some mechanical means which will automatically register the exits and returns of the bees over long periods of time. 1 This paper was prepared in part fulfillment of the requirements for the degree of doctor of philosophy at Cornell University. While engaged in this work the author held a Government overseas scholarship from the Department of Agriculture of the Union of South Africa. The work was done in cooperation with the Bee Culture Laboratory of the Bureau of Entomology, at Somerset, Md. The writer isi indebted to the Carnegie Fund of the Imperial Bureau of Entomology, London, and to the Department, of Agricul- ture, Union of South Africa, for financial assistance. He was regularly assisted in the taking of data by Miss Effle Ross and from time to time by other members of the Bee Culture staff. lie also gladly acknowledges his indebtedness to Prof. F. Y. Edgeworth, All Souls College, Oxford, several of whose published articles have been found of value, and to Prof. H. FI. Love, Cornell University, and Dr. G. W. Vinal, Bureau of Standards, for helpful suggestions. 26909°— 25 1 2 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE /,]'■■ :.i i/, .',,.31 . While reading an interesting paper by Leon Dufour,2 in which are shown the great variations which the weight of a hive under- goes in the course of a single day, the writer, seeing how the in- coming nectar and pollen modified the interpretations of the flights which occurred, realized more fully the scientific value of an automatic apparatus which wTould register the exits and returns of the individual bees, and was led to spend the greater part of the winter of 1921-22 in devising an apparatus for this purpose. Actual recording of individual flights was begun on April 8, 1922, with 14 units of the apparatus in place. On May 10 the full quota of 30 instruments was installed and readings were continued regu- larly, including Sundays and holidays, from daylight to dark, until July 29, except for six days, when the work was interrupted by certain necessary modifications or adjustments of the apparatus. During this period some five million flights to and from the hive were recorded, the gross weight of the bees representing half a ton. Since the first object of this experiment was to determine the prac- ticability of obtaining data on problems pertaining to the flight of bees by means of an automatic recording mechanism, rather than to make a study of any one of these problems, and since there appears to be no immediate opportunity for the writer to continue this in- teresting line of investigation, it seems best to record the data so far obtained, that they may serve as at least an introduction to a further study of this phase of bee behavior. THE APPARATUS At the outset it was clear that, on account of the limited space at the hive entrance, the most practical apparatus for counting the flights of bees would be one in which each bee would establish elec- trical contact as it left or as it entered the hive. Any such appa- ratus must consist of as many units as will accommodate the full flight of the colony under experimentation, each unit consisting of one contact device and one recording mechanism. THE CONTACT DEVICE A device for establishing electrical contact's that will give the de- sired data is one which requires for its operation little or no exertion on the part of the bee and which will delay the bee very little or not at all in its departure or return. To meet these requirements there are at least three possible methods of approach, arranged here in the ascending order of their complexity and expense. (1) To allow the bee to push against some mechanism. (2) To allow the weight of the bee to operate some mechanism. (3) To use a minute electric current, the circuit of which is closed by the body of the bee itself as it passes the terminals, but not strong enough to cause any modification of behavior, and then to amplify this current so as to operate the recording mechanism. Working on the plan of allowing the bee to push against some mechanism, a modification of the ordinary bee-escape which allows bees to pass but one way was tried. A small wire was soldered at 2 Dufour, Leon. Travail des butineuses et r<5eolte du mid. In L’Apiculteur, vol. 41, no-. 8, pp. 300-312. 1897. THE FLIGHT ACTIVITIES OF THE HONEYBEE 3 right angles to one of the springs, so that as the bee passed between the springs this wire was pushed into a small cup of mercury, thus closing the electric circuit. It is a simple enough matter to get a bee to form such a contact, but to get each contact to represent the passage of but one bee proved to be exceedingly difficult and is per- haps impossible. Seven different devices of this sort were con- structed and tested, each intended to overcome some defect inherent in those preceding. Regarding these devices, it suffices to say that none of them proved practical. One of the chief difficulties was the fact that the springs may be so delicately adjusted that on the passage of a bee a contact may be formed by the head and thorax and then another contact formed by the abdomen. Yet another bee passing these same springs in a slightly different manner may get through without forming any electrical contact whatever. An at- tempt was made to overcome this difficulty by taking advantage of the smaller constriction which exists between the dorsal surface of the propodeum and the abdomen ; that is to say, the bee was forced to walk in an upright posture, through a tube of special cross sec- tion, so that the spring passed over this dorsal region. The bees, however, showed great reluctance to go through such a tube. George S. Demuth and N. E. Mclncloo, both formerly con- nected with the Bee Culture Laboratory, have informed the writer that they, too, have given some thought to this problem. Mr. Demuth attempted to get the bees to push against prongs placed equidistant around a revolving wheel, much on the principle of the undershot water wheel. Doctor Mclndoo tried to get the bees to push against small hinged gates. Neither attempt was carried to completion. Owing to the fact that the stimulations of bees to flight, such as light intensity, nectar flows, and other environmental factors, vary greatly in degree, and also that individual bees apparently differ somewhat in their reactions to even slight obstacles placed in their way, it is questionable whether any apparatus working on the prin- ciple of having the bees push against a mechanism will ever be found practical. Attention was then turned to the possibility of utilizing the weight of the bee to form an electrical contact. Three models were con- structed on this plan, each proving to be a step in advance. The third model, after several long tests, appeared to be the instrument required. After a few minor improvements in the design, 30 such instruments were built, 15 to accommodate outgoing bees and 15 for incoming. These instruments are referred to in this paper as “ gates.” This device (fig. 1) may be described as a miniature balance on jeweled pivots. As the bee enters the tunnel fixed to one end of the lever, its weight, having a greater moment than the counterbalance, causes the tunnel to drop and this movement produces three con- secutive results: (1) It closes the door to prevent a second bee from gaining entrance to the case while the tunnel is on its downward stroke; (2) it establishes an electric contact by thrusting two plati- num prongs attached to but insulated from the lever, into two mercury cups; and (3) it opens a. second, door on a lower level, per- mitting the bee to fly to the field or to enter the hive, as the case may 4 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE be. The bee having left the tunnel, the counterbalance now has a greater moment than the empty tunnel, the second door is closed, contact is broken, and the first door is opened, so that the mechanism is now ready to count the next bee. This seesaw motion continues as long as bees are passing through the tunnel. The illustration (fig. 1) shows an incoming gate. The only difference between this and an outgoing gate is the presence of glass immediately in front of the tunnel on the outgoing gate, and a different position of the binding posts. The tunnels adopted after some experimentation were 15 millimeters long, from 6.5 to 7 millimeters wide, with a curved upper portion from 4.5 to 5 millimeters at the highest point. These were so adjusted that a 67-milligram weight placed at the base of the rear door brought the tunnel down to the lower stop. At other times 75-milligram and 47-milligram weights were used, but most of the data were obtained with the first-mentioned adjustment. Fig. 1. — Apparatus for recording by electrical contact the ingress of bees into the hive Since this device, in the course of this investigation, has shown certain undesirable features, it remains, in a search for the ideal counting device, to consider briefly the possibility of counting bees by means of the amplification of a current whose circuit is closed by the body of the bee itself. In discussing the possibility of this method with the writer, radio experts have suggested that it might be better to get the bee to pass between the plates of a small condenser and thus to vary its dielectric constant, and then to use this variation to record the passage of the bee. Assuming that expense is of no consideration, it is still ques- tionable whether this method would have any advantage over the one used in this investigation. Some device would be needed before and after the condenser to regulate the passage of the bees, and these obstructions might be as great as or greater than that offered by the THE FLIGHT ACTIVITIES OF THE HONEYBEE 5 present device, so that the only advantages to be gained, provided the principle could be developed to do away with occasional multiple recording by the same bee, would be a less frequent necessity for cleaning the gates, and perhaps a relatively larger capacity for each gate. THE COUNTING DEVICE If the bee-escape method of counting had proved successful, it would have been necessary to use a weak current for making the records. The first type of counter tried consisted of an ordinary alarm clock with the balance wheel removed. A piece of soft iron was then attached to the escape lever and an electromagnet was placed on the framework. When this magnet was excited by the closing of the circuit by the bee, it attracted the soft iron on the escape lever and allowed one tooth of the escape wheel to pass. When contact was broken, a light spring brought the lever back to its first position. Thus one contact corresponded to one vibration of the original balance wheel, and by knowing the beat of the clock it was possible to count the exits or entrances from any particular gate merely by reading the “ time ” on the dial. This instrument, though crude, proved promising, but was not adopted for the present investigation. With the balance device adopted, a much stronger current can be used. It was therefore pos- sible to use a telephone-message register, already manufactured and available, which consists of a simple cyclometer actuated by an electromagnet, operating on a higher voltage (about 16 or 18 volts) than would be required for the counter made from the clock. ARRANGEMENT OF THE APPARATUS For convenience in handling, each set of gates was attached to a board hinged to the end of a 10-frame Langstroth hive body, so that the instruments could be swung away from the hive and yet not be detached entirely. When these boards were clamped to the hive, so as to close the entrance, the outgoing gates were above and somewhat in front of the incoming ones, the actual exit apertures being about 2 inches above the row of entrance apertures. This arrangement proved very satisfactory, because it minimized the possibility of the bees holding dowm the tunnels of outgoing gates by attempting to enter the hive through them. A false bottom-board within the hive conducted the bees to the outgoing instruments, and an outside alighting-board led the return- ing bees to the ingoing tunnels. The incoming bees, on passing through the contact mechanism, entered a small chamber below the inside false bottom-board, and to enter the brood chamber proper they passed through slots in this false bottom-board. A strip of queen-excluding zinc was placed behind the outgoing gates and above the slots at an angle of about 45°, sloping upward and back- ward. This forced the bees in carrying their dead to drop them, so that the dead bees fell through the slots in the false bottom-board to a position behind the incoming gates, where they could conveniently be removed and counted. Ventilation was at first provided by means of a maze which admitted air but no light through a slot in the bottom-board. This 6 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE arrangement proved satisfactory until some time in June, when a change was necessitated by the peculiar behavior of the bees during hot weather. Several changes were tried, but the plan that ap- peared to give the greatest satisfaction was to close this lower venti- lator and to remove one of the ingoing gates, the remaining 14 gates being spaced more widely, thus allowing more of the hive odor to escape between the gates. Ventilation at the top of the hive was also provided during the hottest weather. A weatherproof telephone cable led the wires from the contact de- vices to the recording counters attached to a table in the laboratory, about 50 feet from the hive. Two storage batteries furnished the current, these being recharged periodically by means of a rectifier. During the portion of each day when the bees were flying, the following observations were recorded without interruption at exactly 15-minute intervals : Records of outgoing and incoming bees ; weight of the entire hive with attached apparatus ; temperature in the shade adjacent to the hive; anemometer record of the wind which had passed since last reading; and notes on the weather conditions, includ- ing the degree of cloudiness. These readings usually began at about 5.30 a. m. and continued until about 8 p. m., depending somewhat on the season. PROBLEMS ON WHICH INFORMATION MAY BE OBTAINED BY A DEVICE FOR COUNTING FLIGHTS If an accurate count could be obtained of the exits from and en- trances to a hive without interfering unduly with the normal flight activities of the bees, the data obtained would throw some light on the following problems : 1. Perhaps the most important would be the responses of bees to various honey flows, especially to those that are not heavy enough to be reflected very markedly in the weight of the hive. Information regarding this problem would undoubtedly assist in a study of nec- tar secretion, especially with reference to the time of day and the rel- ative amount of secretion. 2. The responses of bees to various meteorological conditions, such as intensity of light, temperature, wind, rain, and electrical disturbances. 3. Average duration of the flight, its variation with the honey flows and the general atmospheric conditions. 4. The daily average number of trips per bee to the field, ascer- tained by manipulating the hive so as to get a census of the field bees on any particular day. 5. The possible responses of bees to sounds, odors, and other stimuli. 6. The death rate of the colony, the comparative numbers which die in the field and in the hive, and the factors which contribute to an increased death rate. 7. The effect on bees of the time of application of certain poisonous insecticides in horticultural practice. 8. The relative economic importance of predatory wasps, other enemies, and adult bee diseases, information regarding which might be obtained by correlating flight data with the normal death rate. THE FLIGHT ACTIVITIES OF THE HONEYBEE 7 THE CONDITIONS OF THE EXPERIMENT In order to care for the great rush of bees on the approach of storms or to allow for possible clogging of some of the gates, it was necessary to provide such conditions that the gates would be worked normally much below their full capacity. This was done by employ- ing a colon}T of bees of comparatively small size. When the taking of the records was begun in April, this colony, headed by a 1-year- old Italian queen, was in a hive composed of two bodies. The col- ony had about 5 pounds of bees, brood to fill four frames completely (Langstroth size) , and seven frames containing honey. The remain- ing frames were nearly empty. When the colony was examined in detail on May 25, seven frames of brood and two frames containing eggs were present. A considerable number of larvae then showed symptoms of sacbrood. On the termination of the experiment at the end of July there were about 3 pounds of adult bees, three frames of brood, and a serious shortage of honey. Some of the larvae dead of sacbrood were not yet cleaned out. To prevent so far as possible the drifting of other bees to this colony, the hive was set up in an isolated position about 36 feet from another colony and about 100 feet from the main yard on the west side of the bee culture laboratory, Somerset, Md. It was shaded by the building until 9 to 10 a. m. and by a tall tree in the afternoon. During April, 1922, the hive did not show any days of gain in weight due to incoming nectar. In this region the main honey flow comes rather early, the two main nectar sources being black locust ( Robinia pseudacacia) and tuliptree ( Liriodendron tulipifera). These species yield nectar in May, usually early in the month, fol- lowed by a small amount of nectar from white clover ( Trifolium repens) and from other plants of minor value for nectar. Usually b}^ the end of June in this region a dearth of nectar occurs and lasts imtil fall, so that there is nothing for the bees to gather during midsummer, unless, as sometimes happens, there is a production of honeydew. In the season of 1922 the first gain in weight from black locust occurred on May 3, but a heavy rain on the night of May 14 brought this honey flow to a sudden close. The tuliptree began to bloom on May 7, and the last gain from this source occurred on May 28. During this period of substantial honey flow there was no gain on five days because of rain. Small gains in weight from in- coming pollen, with some nectar, took place in this colony on June 8, 10, and 17. All the other days throughout June and July showed a daily loss in weight, with the exception of July 16, when in the evening the hive had exactly regained its morning weight. The highest daily gain recorded for this colony was 1.440 kilograms, and on two other days the gain exceeded a kilogram. If a colony of full strength could have been employed the gains in weight would have been larger. Of the 50 days on which this colony showed a loss in weight and on which rains did not invalidate the scale readings by the accumu- lation of moisture on the hive, 9 days showed a loss of 10 to 90 grams, 15 a loss between 100 and 190 grams, 8 between 200 and 290 grams, 13 between 300 and 390 grams, and 4 between 400 and 490 grams, with 1 day showing a loss of 610 grams. The experiment was discontinued before the beginning of the autumn honey flow. 8 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE For convenience in compiling the data, the quarter-hourly read- ings were recorded by means of a typewriter on cards of three differ- ent colors, blue being used for the outgoing data, yellow for the incoming data, and white for all other records and general notes. The differences between the successive readings and other data were obtained at a later date by the use of an adding machine. The long days, with frequent observations and the dismantling and cleaning of some of the gates in the evenings to get the apparatus ready for the following day, not only necessitated two observers but also rendered it impossible to make compilations during the active season of work. FACTORS INTRODUCING ERRORS IN THE COUNT With any untried experimental mechanism consisting, as this did, of many units, each unit in turn possessing parts having defects either of construction or of design, it is too much to expect perfect accuracy during the first season. Before discussing the magnitude of the experimental error produced by these deficiencies, it is desir- able so far as possible to know the manner in which this error arises. 1. After observing the passage of about 2,000 bees through the early experimental model, it was thought that a tunnel had been devised which although allowing only one- bee to pass at a time, would cause little inconvenience to the insect and would avoid as much as possible the scraping off of the load of pollen as the bee passes through the device. However, further experience with the full set of gates on the hive showed that some of the tunnels would occasionally allow two bees to pass at one time. Sometimes the two bees would get through and establish only one electrical contact, but they might cause two or even more contacts because of the irregular manner in which the tunnel fell under their weight, the movement being impeded by the pressure of their bodies against the stationary surfaces of the apparatus. Under normal weather conditions this difficulty was confined almost entirely to the outgoing gates, as the stimulus urging the bees into the outgoing channels was apparently much stronger than that attracting them into the ingoing tunnels. By darkening the glass windows and reducing the dimensions of the tunnels, this error was reduced to a minimum. Rebuilding all these tunnels was a tedious operation, consumed a great deal of time, and could be done only gradually, so that the early records show a greater error from this cause than the later ones. 2. Rebounding of the empty tunnel so as to form a second contact occurred occasionally if the adjustment of the counterweight was too delicate or if it had been rendered so by debris or pollen collecting in the tunnel. 3. Debris dropped or propolis placed by the bees in the clearance between the movable and stationary parts may cause double con- tacts by slowing down the speed of the tunnel in its fall; but when the quantity of debris became so large that the tunnel ceased to function, no error was introduced. 4. The voltage made necessary by the adoption of the telephone- message register unfortunately caused a considerable amount of oxi- dation on the surface of the mercury through arcing. In spite of frequent cleaning of the surfaces of mercury, this occasionally gave THE FLIGHT ACTIVITIES OF THE HONEYBEE 9 some trouble in case the mercuric oxide collected as a flocculent mass on the platinum points. Under such circumstances the weight of the bee is no longer sufficient to overcome the greater displacement of mercury now necessary, and multiple contacts may result. This error was apparently eliminated by adjusting the apparatus so as to bring the mercury cups closer to the fulcrum, giving the bee a greater moment than this opposing force. 5. The bees about to enter the tunnel may push against the closed inner door (especially if its clearance is large) and may occasionally produce multiple contacts when the tunnel is either on its downward or on its upward stroke. 6. A bee pushing hard against the glass of an outgoing gate may produce multiple contacts by preventing a tunnel from falling rapidly past its critical point. Errors from this cause and from pushing against the closed inner door were very much reduced by darkening the glass on the gates (p. 8). 7. Bees clawing at the closed outer door of the device may bring the tunnel down far enough to form an electrical contact, especially if the doors are roughened by dirt or propolis or by a corrosion of the surface of the metal. Error from this source was peculiar to the ingoing gates, for on only two occasions was it observed on an out- going gate. An attempt was made to eliminate this source of error by placing small metal cups over the exit hole so that clawing could take place only in an upward direction. 8. The segments of some of the recording cyclometers would some- times bind against each other, thus failing to record the contacts when the magnet was excited by the passage of a bee. This necessi- tated some readjustments of the cyclometers. 9. The cleaning of the gates in the evening, after the day’s records had been taken, necessitated the use of an artificial light, which on warm nights attracted some bees from inside the hive. These might remain outside all night, thus introducing an error on the following day. 10. The clustering out of the bees in the warm weather following the main honey flow was one of the most disturbing features of this investigation. On many days, records which were normal during the early part of the day were rendered valueless later on through the clustering out of the bees in the afternoon. At the same time this clustering out occurred throughout the apiary. 11. In the construction of the outgoing gates a small strip of metal was left between the outer aperture and the lower edge of the glass window, which was sometimes grasped by the bee in its fall, giving rise to multiple contacts. 12. A small error was caused by the actual drifting of the bees. This was apparent on colder mornings when the other hives in the apiary were active and the experimental one had not yet commenced its activity. It is clear from the preceding list of causes of errors that with the exception of No. 8, and sometimes of No. 1, all these factors have a tendency to increase rather than decrease the recorded exits and entrances of the bees. With the experience gained in the 1922 season in the design and handling of this apparatus, the writer is convinced that, with the exception of Nos. 2 and 3, where debris is the cause of error, all these factors may eventually be eliminated. 26900° — 25 2 10 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE On any particular day the magnitude of the experimental error depends upon the combination of factors operating on that day. On 89 of the 105 days on which records were taken, omitting records accidentally falsified by various causes, it appears that out of every 100 of the 2,484,666 contacts formed by outgoing bees, 96.84 re- turns were registered. On 37 of these days the records show more bees coming back than were recorded as going out, and these so- called gains in the returns give an obvious error of at least 1.86 per cent for the days recorded in Table 1. These 37 days were dis- tributed as follows, and showed for every 100 exits daily the fol- lowing average returns for each group : April, 4 days, 111.99 ; May, 10 days, 104.50; June, 7 days, 104.87; July, 16 days, 104.29. These data give some idea of the magnitude of the daily experimental error which may take place. The percentage returns on these days varied from 100.08 to the maximum, one day in April, when they were as high as 127.81. If the days are selected on which the apparatus worked well, and the daily percentage of error which occurred, so far as it could be determined, is kept in mind, the general utility of the curves ob- tained from these data for such purposes as a study of the effect of external environmental factors is little affected by this error. For a study of the average duration of each flight, where a greater degree of accuracy is essential, a closer selection of data is necessary. Referring, therefore, to the outline of the problems on which it is hoped that such an apparatus may give some information, it is found that this apparatus is performing its function as regards four of them, one of which has been rated as the most important in this in- vestigation. FACTORS INFLUENCING THE FLIGHT The various activities of the colony population are so interrelated that in a study of any one factor, influencing any particular activity such as flight, it sometimes becomes exceedingly difficult to gauge its exact influence, or even to give the right factor the credit for the behavior observed. Since it is possible not only to observe the varia- tions in flight produced by changes in the intensity of any single factor throughout the course of a day, but also to gauge to a certain extent the gross influence of this factor on the day’s flight as a whole, it is necessary in a study of any factor to make use of both of these sources of information. To ascertain the gross influence of any factor, a general survey of the whole period in which records were made is necessary. THE SEASONAL SURVEY In order to obtain a comprehensive picture of the daily flight activities throughout the season, it is necessary to plot a curve (fig. 2) of the total daily exits from the hive (Table 1). As might be expected, this curve has several high points, representing the data for those days when the conditions for flight were at an optimum for the period of the season in which they occur. A day that is consid- ered an optimum for April would very naturally be a bad flight day for June or July; therefore in a study of any particular day a com- parison must be made between this day and another in close prox- imity to it, when presumably, or as far as can be ascertained, the field conditions, internal conditions of the colony, and the number of THE ELIGHT ACTIVITIES OF THE HONEYBEE 11 field bees are but little different. By conecting these high points in the flight curve, there i§ obtained a second curve (fig. 2) indicat- ing the optimum seasonal flight possible. This optimum curve, s9/=>je/t. amy Sixteen specimens reared in Tennessee determined by W. R. Walton. 18 BULLETIN 1336, U. S. DEPARTMENT OF AGRICULTURE Winthemia quadripustulata Fab.14 was reared by F. M. Moody at Charleston, Mo., and by the writer at Knoxville, Tenn., in 1916. The caterpillars from which the latter were reared each bore two small, oval white eggs on its thorax ; and, when swept from the field, one was a prepupa, while the other was still feeding. The former pupated a few days after capture, and when examined three days later a large puparium was found filling over two-thirds of its interior, leaving only the end of the abdomen empty. The other caterpillar also pupated; but in this case the parasite larva, a yellowish- white maggot 8 millimeters long, emerged from the pupa and formed its puparium outside. In addition to the foregoing Diptera, Sherman (10) listed as reared in North Carolina the following five species: Bombyliidae: Anthrax lateralis Say. Tachinidae: Euphorocera Jloridensis Tns., Exorista boarmiae Coq., Frontina aletiae Riley. Sarcophagidae : Sarcophaga cimbicis Tns. HEMIPTERA NABIDAE Nobis ferns L.15 — This slender gray bug has been found in the field at different times feeding on the young Plathypena scabra. As it has usually been found exceedingly abundant in infested fields in Tennes- see examined by the writer, it undoubtedly aids considerably in the destruction of the caterpillars. The nymphs as early as the first and second instars have been observed to attack and kill the young arvae. Bugs kept in captivity deposited eggs in rows along the stems of the alfalfa plants, each egg inserted deeply, with only one end showing on the surface as a tiny white spot. PENTATOMIDAE Podisus maculiventris Say, the spined soldier-bug, has been found numerous in infested fields and undoubtedly kills many of the cater- pillars. One bug was found in the field with a Plathypena scabra larva pierced by its beak. This specimen was determined as Podisus maculiventris Say by O. Heidemann. Individuals kept in captivity fed readily on the caterpillars, one destroying five in the course of five days’ captivity. It pierced the larva with its beak and sucked its contents, leaving only a shrunken remnant of skin and solid parts. FUNGOUS DISEASE In the fall of the year great numbers of larvae are killed off by the fungus Botrytis rileyi Farl.16 Both at Knoxville and Nashville, Tenn., they have been severely attacked by this disease, and similar reports have come from Hagerstown, Md. 14 Three specimens from Tennessee determined by W. R. Walton. 15 An individual found by the writer feeding on a Plathypena scabra larva was identified by Herbert Osborn as Nabis ferns L. 46 Infestation on Plathypena scabra larv* collected by the writer at Knoxville, Tenn., was determined as this fungus by Alden T. Speare. BIOLOGICAL STUDIES' OF THE GREEN CLOVER WORM 19 LITERATURE CITED (1) Barnes, W., and McDunnough, J. 1917. Check list of the Lepidoptera of boreal America. ix, 392 pp. Decatur, 111. Synonymy of Plathypena scabra, p. 92. (2) Chittenden, F. H. 1901. Some miscellaneous residts of the work of the Division of Entom- ology, V. U. S. Dept. Agr., Div. Ent. Bui. 30, n. s., 98 pp. The green clover worm, pp. 45-50, fig. 26. General account, with descriptions and figures of egg, larva, and adult. (3) Grote, A. R. 1873. Conclusion drawn from a study of the genera Hypena and Her- minia. In Buffalo Soc. Nat. Sci., vol. 1, pp. 37-40. The name Plathypena scabra is proposed, p. 38. (4) Hawley, I. M. 1922. Insects and other animal pests injurious to field beans in New York. Cornell Univ. Agr. Exp. Sta. Mem. 55, pp. 949-1037. A general account, with record of two parasites, pp. 1011-1014, figs. 96, 97. (5) Hill, C. C. 1918. Control of the green clover worm in alfalfa fields. U .S. Dept. Agr. Farmers’ Bui. 982, 7 pp. General account and control measures in alfalfa. (6) Howard, L. O. 1885. Descriptions of North America Chalcididae from the collections of the U. S. Department of Agriculture and of Dr. C. V. Riley, with biological notes . . . U. S. Dept. Agr., Div. Ent. (o. s.) Bui. 5, 47 pp. Euplectrus platyhypenae, n. sp., pp. 26-27. (7) Lintner, J. A. 1873. Hypena scabra (Fabr.) and H. erectalis, Guen. In Canad. Ent., vol. 5, pp. 81-82. Male and female forms identified as the same species. (8) Muesebeck, C. F. W. 1922. A revision of the North American ichneumon-flies belonging to the subfamilies Neoneurinae and Microgasterinae. In Proc. U. S. Nat. Mus., vol. 61, art. 15, 76 pp., 1 pi., sep. no. 2436. Two parasites recorded reared from Plathypena scabra, pp. 32, 66. (9) Riley, C. V. 1880. The cotton worm. U. S. Ent. Comm. Bui. 3, 144 pp. Records hibernation of adult and pupa of Hypena scabralis (Fabr.), with note on food plants. (10) Sherman, F. 1920. The green clover worm as a pest on soy Deans. ±n Jour. Econ. Ent., vol. 13, no. 3, pp. 295-303. A general account of its occurrence in North Carolina, with measures for control. (11) Smith, J. B. 1895. A revision of the deltoid moths. U. S. Nat. Mus. Bui. 48, 129 pp., 14 pis. (Contributions toward a monograph of the insects of the lepidopterous family Noctuidae of boreal North America.) Description of genus and species, with figures of moth, pp. 110-112, pi. 9, nos. 10, 11, 12. ORGANIZATION OF THE UNITED STATES DEPARTMENT OF AGRICULTURE August 14, 1925 Secretary of Agriculture Assistant Secretary Director of Scientific Work Director of Regulatory Work Director of Extension IVork Director of Information Director of Personnel and Business Adminis- tration Solicitor Weather Bureau Bureau of Agricultural Economics Bureau of Animal Industry Bureau of Plant Industry Forest Service Bureau of Chemistry Bureau of Soils Bureau of Entomology Bureau of Biological Survey Bureau of Public Roads Bureau of Home Economics Bureau of Dairying Fixed Nitrogen Research Laboratory Office of Experiment Stations Office of Cooperative. Extension Work Library Federal Horticultural Board Insecticide and Fungicide Board Packers and Stockyards Administration Grain Futures Administration W. M. Jardine. R. W. Dunlap. Walter G. Campbell. C. W. Warburton. Nelson Antrim Crawford. W. W. Stockbergek. R. W. Williams. Charles F. Marvin, Chief. Henry C. Taylor, Chief. John R. Mohler, Chief. William A. Taylor, Chief. W. B. Greeley, Chief. C.’ A. Browne, Chief. Milton Whitney, Chief. L. O. Howard, Chief. E. W. Nelson, Chief. Thomas H. MacDonald, Chief. Louise Stanley, Chief. C. W. Larson, Chief. F. G. Cottrell, Director. E. W. Allen, Chief. C. B. Smith, Chief. Claribel R. Barnett, Librarian. C. L. Marl att, Chair man. J. K. Haywood, Chairman. John T. Caine, in Charge. J. W. T. Duvel, Acting in Charge. This bulletin is a contribution from • Bureau of Entomology L. O. Howard. Chief. Division of Cereal and Forage Insect In- G. A. Dean, Senior Entomologist, vestigations. in Charge. 20 ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT ,r. CENTS PER COPY V UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1339 Washington, D. C. Y October, 1925 THE EFFECT OF WEATHER UPON THE CHANGE IN WEIGHT OF A COLONY OF BEES DURING THE HONEY FLOW By James I. Hambleton, Apiculturist, in Charge of Bee Culture Investigations, Bureau of Entomology CONTENTS Page Introduction - 1 Method of obtaining data 5 Method of presenting data ... 8 Comparison of changes in weight of two colo- nies of bees 11 The spring period 15 The fall period 22 Correlations between external factors and the changes in colony weight 25 Page The effect of unknown factors on changes in colony weight 38 Theoretically changing weather factors and pre- dicting resulting gains 44 Conclusions 48 Literature cited 50 INTRODUCTION The annual honey crop of a colony of honeybees is dependent upon a considerable number of factors, part dealing with the activities of the bees and part resulting from the various external factors influ- encing the secretion of nectar by the honey plants of the locality. Demuth (8, y. 18 y has shown this interrelationship by pointing out that there are four factors which combine to make the honey crop : A surplus population in the colony over and above the bees neces- sary for colony maintenance, a predominance of the storing instinct and the control of swarming, honey plants growing under opti- mum conditions, and weather suitable for the secretion of nectar and the gathering of it by the bees. If any one of these factors is reduced to zero, the crop is zero, and if any one factor is reduced one-half, the erop is one-half the maximum. Naturally, so long as the factors do not rest on a mathematical expression, it is impossible to inter- pret them with exactness, but every experienced beekeeper realizes this interrelationship; it is therefore sale, as a working hypothesis, to accept these factors as real and fundamental. Most of the work done on beekeeping subjects has dealt either with methods of obtaining for the colony a surplus population at the right time for the gathering of the crop, or with the management ol 1 Kcfercnce is made by number (italic) to “ Literature cited, ” p. 50. 42201—25 1 2 BULLETIN 1339, U. S, DEPARTMENT OF AGRICULTURE the colonies so as to induce the bees to expend their energy in gath- ering, which, as every experienced beekeeper knows, means the con- trol of swarming and any other instinctive activity which might tend to reduce the manifestation of the gathering instinct. The last two factors of those above mentioned have to a considerable degree been neglected, doubtless because they are outside the control of the bee- keeper in any given location. Since the honey crop is so intimately connected with these factors, however, it is unwise to neglect them. There are extensive records and lists of the plants which furnish nec • tar in quantities sufficient to make beekeeping profitable in the var- ious parts of the country, and there are certain results of botanical investigations which bear on this subject, but as a rule these results have not been part of the beekeeping literature. Almost no atten- tion has been paid to the effect of weather factors on the gathering of the crop. The purpose of this bulletin is to present information on the rela- tionship existing between changes in the weight of a colony of bees during a honey flow and the prevailing weather conditions, based on data obtained at the Bee Culture Laboratory, Somerset, Md., from February to November, 1922, and for the month of May, 1923. The major problem during this time was an intensive study of colony temperatures, but this experiment necessitated the recording of changes in the weight of at least one colony of bees hourly throughout the day and night for long consecutive periods. At the conclusion of the experiment a hasty survey of these changes immediately re- vealed an abundance of interesting data which seemed to throw con- siderable light upon the relationship of changes in hive weights to outside conditions. Although the data herein recorded on the fac- tors influencing changes in weight are not as complete as they should be to carry such a problem to a final conclusion, the subject is here presented from the material available in the hope that it will serve as a stimulus to investigators in different great honey-producing areas to study the relationship of weather to honey production. A problem like this can not be solved without such comparative data. Information of this sort covering the principal honey-producing sec- tions of the United States would be of inestimable value in the fur- therance of beekeeping. A knowledge of existing honey flora and a correlation of weather conditions with bee behavior should help the prospective beekeeper in choosing the best beekeeping locality. At the present time there is no method of predicting whether a locality will prove profitable to the beekeeper except by the results obtained by other beekeepers. It would be well to know just why certain plants produce nectar in one locality and not in another, and, more important still, to discover if possible the laws underlying the rela- tionship between bees, nectar-producing plants, and weather condi- tions as they influence the results of the beekeeper, and to determine, if possible, which are the most important factors upon which the honey crop depends. In this discussion the results of the relationship of bees and nectar- producing plants are shown either as surplus honey or as increase in the weight of the colony. The two results are inseparable as far as the present work is concerned. No effort was made to study plant behavior and bee behavior as separate subjects, but the results WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 3 7-/A/£T OS' £>S?y' y 8 & /o // /? 7 £ a <2 s’ 6 y 8 & of the two working together as one, and recorded as “colony weight changes/’ hare been studied in their relation to the prevailing weather conditions. It is a very common practice among the best beekeepers to main- tain what is known as a scale colony. A colony of average strength is usually chosen for this purpose and is placed on scales in the apiary. This colony is weighed once or twice a day, usually in early morning and after the bees have stopped flying in the evening. Such weighings however, are usually made in a haphazard manner, with- out much care that the weighings are recorded at an exact time each day. Such records give little information except that they indicate in a rough manner the trend of the honey flow. Unfortunately, even such records as these are not available for most localities. Records of careful hourly weighings of colonies are scarce. Dufour (10) pointed out the value of hourly weighings and showed how enormously the hourly changes in weights throughout the day may vary on days showing the same net gain. Figure 1, taken from data used in the present investigation, illustrates graphically the activities on two days when the net gains were approxi- mately the same. The times of regaining the original weights are shown at the intersec- tions of the curves with the zero line. Dufour (10) gives information on the weather condi- tions only in a general way, during the time when he carried on his experiments, and it is therefore impossible to calculate the exact relationship in his investi- gation between the changes of colony weight and the weather. In any discussion of changes in colony weight it is important to keep in mind the fact that these changes are brought about by two factors working jointly, both influenced by weather conditions. These factors are the secretion of nectar by the plants and its collec- tion by the bees. Of the two factors, investigators have given more attention to nectar secretion than they have to bee behavior. Even in the field of nectar secrel ion there is no general agreement as to the - 4-00 - 800 —/200 -/600 — 1 — 1 i i i i 1 1 1/ ! \ two days when the net gains were approximately the same. The solid line presents the changes on May 9, the other those on Sep- tember 28, 1922 4 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE relative value of the influencing factors; certainly none based on a mathematical determination of their relative importance. Kenyon {21) and Kenoyer {20) called attention to Dufour’s work in the hope that others would carry on similar experiments, but no extensive research seems to have been undertaken until the present work. Discussions of factors influencing nectar secretion alone are often somewhat misleading to the beekeeper, because factors which may stimulate or retard nectar secretion may not be apparent at the hive as shown by changes in weight or in the honey crop. Considerable work has been done on the problem of nectar secre- tion and the influence upon it of temperature, humidity, altitude, latitude, and similar factors, but little is on record other than casual observations by beekeepers as to the influence of these factors on bee behavior during the active season, either in or out of the hive. The works on nectar secretion by Bonnier, Behrens, deLayens, and Ken- oyer stand out prominently in this line. These writers and others have made important contributions on the mechanism of nectar secretion and on the factors influencing it. Unfortunately, only a comparatively few plants have been studied for nectar secretion from a physiological point of view, and when thus studied the- number of individuals of a species has of necessity been small. Too often obser- vations have been confined to flowers cut from the plant in order to control the governing factors, such as temperature and humidity, thus leading to erroneous conclusions. Loftfield {22, p. 101 ) has pointed out in the study of stomatal movement the great difference in behavior between cut stems and potted or field plants of the same species. It is quite evident that the factors influencing nectar secretion are not necessarily synonymous with those affecting the changes in weight of a colony of bees during the honey flow. In glancing over papers dealing with nectar secretion there is usually no clear dis- tinction drawn between the amount of nectar secreted and the sur- plus honey gathered by bees, and one might assume that what affects one should affect the other in like manner. This appears not to be the case, as indicated by the data at hand. It is also evident that data on nectar secretion for certain species of plants should be limited largely to the species in question and not used for comparison with other species in other localities. If climatic conditions affect the behavior of bees in the same manner as they affect nectar secretion, one would expect to find a close correlation between factors reputed to be favorable both to nectar secretion and to increases in the weight of colonies of bees, but such a correlation does not seem to exist. It seems, therefore, either that the proper combination of factors influencing nectar secretion has not been discovered, or that the effect of weather conditions upon the behavior of bees is entirely different from their effect upon plants. It has previously been im- possible to determine by judging from the changes in the weight of a colony of bees whether weather conditions influence the more greatly bee behavior or nectar secretion; and, from the standpoint of the practical beekeeper, the influence of weather upon colony weight is far more important than its influence upon either nectar secretion or bee behavior alone. WEATHER AND CHANGE IN WEIGHT OP BEE COLONY 5 METHOD OF OBTAINING DATA From February to November, 1922, continuous records were made of a colony of bees placed on platform scales sensitive to 10 grams. Owing to the nature of the principal problem under investigation, hive temperatures, the hive was left unprotected and certain of the weight records were rendered useless, water remaining on the hive and bottom-board after rains causing too great errors in the weight records. During most of the month of June, 1922, alterations were necessary in the apparatus for recording temperatures in this colony, and weight records are' therefore not available for this month. A standard 10-frame Langstroth hive was used to house this colony. Previous to the honey flow two brood chambers were in use, and a few days before the beginning of the honey flow the colony was manip- ulated to prevent swarming and additional supers were added. The colony was left in this condition until, in June, alterations were made in the temperature-recording apparatus. During May, 1923, two colonies placed side by side, both on equally sensitive scales, were observed for additional data covering another honey flow. Both colonies were in standard 10-frame hives, each having two hive bodies for brood and three, above a queen excluder, as supers. In this case the colonies were given ample room for brood rearing and storage of nectar, and made no attempt to swarm. Pre- cautions were taken to shield these hives from rain, by placing a sloping board over the entrance of each colony in such a manner that rain would not fall on the bottom-board, while there would be free access to the bees. A false roof was suspended over the outer cover of the hive, which caught the rain and shed it beyond the hive walls. A light wooden framework over the two colonies was covered with canvas. In case of a light rain when the bees were still flying the canvas cover was not used, but during periods of hard dashing rains the canvas was dropped so low as to surround the hives com- pletely and prevent any water from collecting on them. Weighings were made hourly throughout the 24 hours of each day, including Sundays and holidays, three persons being assigned to the work on 8-hour shifts. One of the unfortunate circumstances con- nected with working on such a problem at Somerset, Md., is that the main honey flow is exceedingly short (seldom over 10 days to 2 weeks), thus increasing the probable error in all calculations by re- ducing the number of possible observations. The correlation coeffi- cients for the spring honey flow are based on the one colony in 1922, and on the two colonies in 1923. For conclusions relative to the fall honey flow' only the figures for the one colony of 1922 are available. In order to calculate the relationship between changes of weight and weather conditions during the honey flow, only those days show- ing a net gain of 980 grams or over were used for the spring honey flow. The calculations for the fall period include all days from Sep- tember 4 to October 5, irrespective of gains or losses; the two sets of data are therefore not strictly comparable. Hourly thermograph and hygrograph records were made for use in calculating the mean temperatures and relative humidity; the mean of maximum and minimum temperatures was not used. Ilartzell {16) has pointed out the chances of error in endeavoring to correlate 6 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE the average of maximum and minimum temperatures with various biological activities. A rather rough record was kept of sunshine and cloudiness, hut in all calculations the records of the United States Weather Bureau for Washington, D. C., were emploj-ed. There is a distance of 6 miles between the Weather Bureau Station in Washington and the Bee Culture Laboratory, at Somerset, Md., and it was found that only negligible differences existed beween the recorded hours of sunshine for the two places. Records of solar radiation were also available through the kindness of Prof. H. H. Kimball, of the Weather Bureau, these being taken at American University, in Washington, about 1 mile /from the Bee Culture Laboratory. Calculations relating to the effect of wind upon honey produc- tion were not made. The anemometer maintained at the Bee Cul- ture Laboratory was so located for another purpose that its records were not applicable to this problem. The principal source of nectar available to .the bees in this experi- ment was that from tulip tree ( Liriodendron tulipifera). It is an excellent honey plant in this locality, yielding abundantly for about two weeks in early May. Since it blooms so early in the season, it is of the utmost importance that the beekeeper have his colonies strong, otherwise its nectar is wasted with the exception of what little is used for brood rearing. Occasionally black locust ( Robinia pseu- dacacia) furnishes considerable nectar, but this plant is not depend- able in this region. The period of the secretion of black locust coin- cides closely with that of tulip tree. In 1922 the bees worked on, locust actively for several days, while scarcely a locust blossom was seen in 1923. A rainy May, especially rain for the first few weeks of May, spells crop failure for the beekeeper of this region. In this con- nection, attention should be called to the statement of Kenoyer (19) that a rainy May scarcely fails to precede a good honey season in the State of Iowa. This emphasizes the statement made earlier in this bulletin that with the present limited knowledge of the various honey plants a general application should not be made of data secured in a single locality with a certain species of plant. Kenoyer’s work was based on data obtained in the clover region. The coefficients of correlation between the various factors were ealculated from the following usual formula for biometric calculations (7): Probable errors were calculated from the formula — Tj-fTi .67449 PEr = — 7=— ( 1 — r2) =x1(1 -r2) The values of %i were taken from Pearson (29), and the values of 1-r2 from Miner’s Tables (25) 2. 2 All mathematical calculations are based on accepted biometric methods. The writer here records his thanks to Dr. Sewall Wright, of the Bureau of Animal Industry, for his unfailing assistance and advice in this phase of the work. WEATHER AND CHANGE IN WEIGHT OP BEE COLONY 7 It may not be amiss at this place to give, for those not well versed in biometrics, a brief explanation of the terms “ correlation’ ’ and probable error.” The following quotation is taken from Bowley (5, p. 816) : When two quantities are so related that the fluctuations in one are in sympa- thy with fluctuations in the other, so that an increase or decrease of one is found in connection with an increase or decrease (or inversely) of the other, and the greater the magnitude of the changes in the one, the greater the magnitude of changes in the other, the quantities are said to be correlated. We may have either a positive or a negative correlation. When a change in one quantity or variable is accompanied by a direct change in the other the correlation is said to be positive. A perfect positive correlation has the value of 1. When the correlation is less than per- fect it must be written as a decimal of one, such as .75. When the relationship between two quantities or variables is indirect, such as an increase in one being accompanied by a decrease in the other, the correlation is negative. A perfect negative correlation has the value of — 1. Such a relationship less than perfect must also be written as a decimal and is always preceded by a minus sign. Coefficients of correlation state in numerical terms the relationship between two variables. Graphs are useful in showing relationship between two variables, but they do not give numerical correlation values, and it is often difficult from the study of a graph to discover slight relation- ship or entire absence of relationship; this difficulty does not exist in the case of correlation. It is often convenient to think of a coeffi- cient of correlation in terms of percentage ; thus, a correlation written 0.8654 may be read as 86.54 per cent. The probable error is a term applied in biometrics to make correc- tions in a calculation where complete data are lacking. Sample measurements must be made preliminary to a biometric calculation and it is rarely possible to obtain a complete series of samples. For instance, in calculating the correlation between temperature and the change in colony weight we may have the changes in weight occurr- ing at 80°, 81°, 82°, and 84° F., the change at 83° F., being for some reason impossible to secure. The probable error gives the measure of unreliability due to lack of sufficient data. Obviously, the smaller the number of data involved the greater the probable error. The probable error is written with a combined plus ( + ) and minus ( — ) sign (±), and represents the true correlation as falling somewhere, either above or below the calculated correlation, by a difference most probably equal to the value of the probable error. A cor- relation written .7500 ±.0600 indicates that there is an even chance that the true value lies between .7500 + .0600, or .8100, and .7500 — .0600, or .6900. To be significant, the coefficient of correlation should be at least about four times its probable error. When it is less than this the correlation approaches zero in its significance and is of impor- tance primarily as showing whether a relationship is positive or nega- tive. In discussing the probable error Yule (86, p. 811) says: If an error or deviation in, say, a certain proportion p only just exceed the probable error, it is as likely as not to occur in simple sampling; if it exceed twice the probable error (in either direction), it is likely to occur as a deviation of simple sampling about 18 times in 100 trials — or the odds arc about 4.6 to 1 against its occurring at any one trial. For a range of three times the probable error the odds are about 22 to 1, and for a range of four times the probable error 8 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE 142 to 1. Until a deviation exceeds, then, 4 times the probable error, we cannot feel any great confidence that it is likely to be “ significant.” To reduce the labor of calculating these data, in many cases the original data were coded so that all might be included in 10 classes; this was done so that the calculations could be made on sorting and tabulating machines. Because of this grouping many of the coeffi- cients of correlation are smaller than they would have been had the original grouping been retained. In several cases calculations were made by both methods; the difference in results was in some cases as much as 5 per cent. In no case is this difference sufficient to invalidate the results, and, in fact, the results obtained from coding into 10 groups are somewhat safer, since they tend to give a smaller correlation. Many of the coefficients of correlation shown in Tables 2 and 5 have no direct bearing upon the problem, and therefore are not dis- cussed in this bulletin. They are given, however, to the end that partial correlations between any combination of factors may be cal- culated. It will be seen that high correlations may exists between variables having no direct relationship. In Table 2, for example, there is a coefficient of correlation of .5510 between the net gain and the average temperature of the night following. Obviously the tem- perature of the night can not affect the preceding day’s gain; the correlation exists, nevertheless, owing to the combination of influences of the causative factors upon both the net gain and the temperature on the following night. METHOD OF PRESENTING DATA The graph in Figure 2 represents the manner in which the changes in colony weight during the 24 hours are classified to secure tangible and significant figures as a basis for all calculations. A is a fixed point which, on the vertical scale, represents the weight of the colony at 5 a. m. each day in th> month of May. Loss of weight caused by departure of bees for the field begins at about this hour, sometimes earlier and sometimes later, but a weight taken at 5 o’clock constitutes a suitable average, since 4 o’clock is too early and 6 o’clock is too late for such a start. For the fall honey flow the whole graph is shifted one hour later, to adapt it to the shorter days, placing A at 6 o’clock. B is at the point denoting the lowest weight reached during the day and the hour at which this occurs. Although it may occur either in the morning or in the afternoon, the diminution in weight from A to B is called the morning loss, because this loss always pre- cedes the day’s gain in weight. This morning loss is due to the bees leaving the hive in flight. Morning loss may be graphically repre- sented by a vertical line equal on the scale chosen to the diminution in weight, and will throughout this bulletin be designated by this name. C represents in b >th weight and time the point at which the colony regains its original morning weight. D represents in weight and time the turning point at which the colony has ceased to gain and begins to lose weight. The net gain for the day is therefore represented by a vertical line equal according to the scale to the difference in weight between the early morning reading (at A) and the weight reading at D. The latter point has not necessarily a fixed hour. The vertical distance between the weight coordinates of D and of E represents the WEATHER AND CHANGE IN WEIGHT OF BEE COLONY l 9 nocturnal loss due to evaporation and consumption for colony main- tenance. E is, of course, located on the time scale 24 hours after A, and b?comes the starting point (44) for the following day. The only two points which are fixed with regard to time are therefore A and E. In all diagrams except Figure 2 the base line is drawn, not from the point denoting the weight at the starting point {A), but from a zero point raised by an amount equal to the loss in weight between 4 and 5 (or, in the fall, 5 and 6) o’ lock. This is done to show more clearly in the graphs the time when loss of weight of the hive actually begins, as the bees begin flying in the morning. Calculations of all data were, however, based on zero at the point A (actual weight at 5 or 6 o’clock). Nowhere in the calculations can allowance b3 made for consumption for colony maintenance, which remains an unknown factor. The gai ;s during the day would be more, and the evapora- tion loss at night w mid be less, if this factor could be known. The name “ midday decline ” is given to the decline in the amount ofjgain from hour to hour occurring near midday. Although gains are actually taking place during this period on most days, there is a noticeable difference in the rate of gain at this time as compared with that of the hours immediately before and after. The 24 hours of the day are 'divided into diurnal and nocturnal periods, and the designations a. m. and p. m. are not used, since they have no biological significance. The differentiation of diurnal from nocturnal in this discussion is dependent upon the activity of the bees and not on light or darkness. The diurnal period ends and the nocturnal period commences when the weight of the colony ceases gaining, toward the close of the day, and the diurnal period begins, not necessarily at dawn, but at the hour when bee activity outside the hive becomes noticeable. Obviously, flight in the afternoons may be prevented or reduced by rain or inclement weather; and, to prevent the necessity of discarding data obtained on such days, the point D is located in the calculation of the data for 1922 at the time when bees cease gaining and begin to lose on the days immediately preceding and following the days in question, 42201—25 2 (3/&PAS& 10 BULLETIN 1339. U. S. DEPARTMENT OF AGRICULTURE 77/wf; Fig. 3. — Graphs indicating changes of weight in colony AB, May 22, 1922. The solid line shows the cumulative weights; the shaded areas the variations in weight from hour to hour WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 11 thus making the diurnal and nocturnal periods of approximately equal length for all days studied for that year. This change was necessitated by the errors in weight due to water standing on the hives that year, as previously explained. In calculating the data for 1923 such a shifting of D was unnecessary, since the two hives used that year were under cover and no water stood on them. When rain interfered with flight during midday, and flight was later resumed, the point D takes its natural place at the close of flight activity for the day. A graphic representation of changes in weight for a period of 24 hours is given in Figure 3. Either of two methods may be employed, both being shown in this graph. The heavy dark line represents cumulative gams and losses m weight, while the shaded portion shows differences in weight from hour to hour. In case of the shaded portion that part above the base line (shown as A F in fig. 2) is increase in weight and that below is decrease. It is readily seen that the shaded portion of the graph is more important than the line showing cumulative gain, since it magnifies small changes in the rate of gain or loss which might otherwise not be observed. For this reason the method showing differences in weight from hour to hour has been used in all graphs except those showing the net gain. In certain graphs both methods are used for greater clearness of differ- ences. The effects of the various weather factors upon the hive- weight changes in the spring and fall are so different that these two periods must be considered separately. COMPARISON OF CHANGES IN WEIGHT OF TWO COLONIES OF BEES Beekeepers often assume great differences in the gathering ability of colonies of bees in the same apiary and under exactly similar ex- ternal conditions. In the interpretation of the data obtained in this investigation it might be assumed that the colonies used were either unusually good or unusually poor at gathering nectar and pollen. In order to show that in the following calculations the individual char- acteristics of the colonies play a very minor role, it seems best at this point to insert a correlation of the changes in weight of the two col- onies used in 1923. There is, in fact, little reason to believe that such differences in colonies as have been assumed by many beekeepers actually are important in considering the differences in honey crops obtained by various colonies in an apiary, and this is especially tbe case when observations are confined to a single race of bees, as was true in this experiment. When differences in the total acquisition of adjacent colonies of bees are noted, they must in most cases be attributed to differences in manipulation or care of the colonies, or to tendencies to retard gathering in certain cases because of crowd- ing, or to a dominance of the swarming instinct, rather than to propen- sities for heavy or light gathering by the individual bees. Obviously, in the case of poor queens which are unable to keep up the population of their colonies there will be a reduction in the accumulation of stores 12 BULLETIN" 1339, U. S. DEPARTMENT OF AGRICULTURE of the colonies headed by such queens, merely because of lack of suffi- cient bees ; but this does not indicate any reduction in the propensities to gather of the individual bees. Merrill (2/f) reaches the same conclu- sion in his study of changes in colony weight. Whatever slight dif- ferences occur in the propensity of individual bees or colonies of bees to gather nectar and ripen honey are presumably due to actual anatom- ical differences rather than to marked differences in instinctive activ- ities. Table 1 gives the hourly changes in weight of colonies 1 and 2 for 13 consecutive da}7s in 1923. The two colonies respond to external stimuli with remarkable similarity hour by hour. Every break in 77M£: O/^ AA7 Fig. 4. — Hourly changes in weight of colonies 1 and 2, May 21 and 22, 1923. The shaded portion represents the gains and losses of colony 1 superimposed on those of colony 2. The black areas show the excess of the gain or loss of colony 2 over that of colony 1. (From Table 1.) the weight graph of one colony is almost exactly duplicated in that of the other. Figure 4 represents graphically the weights of colony 1 superimposed on those of colony 2 for May 1 and May 22. The hourly differences in weight between the two are practically identical. So far as the amount of gain is concerned, colony 2 is the stronger, but the two behave almost the same from hour to hour. Figure 5 rep- resents the average changes in weight hour by hour for the three colonies (one in 1922 and two in 1923) during the Ma}7" honey flow. The similarity of these changes in weight of the three colonies is strikingly apparent, despite the fact that the data of colony AB were collected in May, 1922, and those of colonies 1 and 2 in May, 1923. WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 13 Table. 1. — Hourly changes in weight, in grams, colonies 1 and 2, May IS to 30, 1923 Date Col- ony num- ber 5 a. m. 6 a. m. 7 a. m. 8 a. m. 9 a. m. 10 a. m. 11 a. m. 12 m. 1 p. m. 2 p. m. 3 p. m. 4 p. m. May 18 1 -20 -40 -150 140 250 270 290 230 50 330 150 400 Do 2 -40 -30 -180 -160 390 420 360 300 80 490 280 560 May 19 1 -60 -30 -80 -200 -130 80 320 280 170 210 370 350 Do 2 -40 -40 -110 -210 -190 140 330 330 130 340 480 420 May 20- _ 1 -70 -60 -90 -180 0 250 280 430 310 220 310 760 Do 2 -80 -100 -100 -190 90 300 300 460 590 80 460 940 May 21 1 -70 -70 -50 -80 -280 -40 380 370 400 270 330 770 Do 2 -70 -70 -50 -90 -280 -30 460 440 460 270 530 920 May 22 1 -80 -80 -70 -160 -200 100 90 340 350 220 260 47o Do 2 -90 -90 -80 -170 -250 150 220 370 390 300 320 540 Mav 23 1 -50 -60 -70 -80 -220 230 -90 280 390 320 -60 -10 Do 2 -60 -70 -80 -80 -230 300 90 330 370 340 -80 -90 May 24 i -30 -30 -50 -70 -110 140 360 330 0 400 430 450 Do o -40 -40 -60 -50 0 150 510 370 80 430 510 510 May 25 i -60 -30 -60 -120 170 450 340 370 120 140 250 480 Do 2 -60 -50 -60 0 150 490 400 430 140 290 330 700 May 26 1 -50 -60 -80 50 170 330 530 290 120 30 240 500 Do 2 -60 -80 -50 140 190 330 450 380 60 190 450 660 May 27. 1 -50 -50 -20 220 230 350 370 330 360 100 290 480 Do 2 -70 -70 50 340 200 340 410 370 240 30 420 560 May 28 1 -60 -50 -50 -60 20 130 270 300 520 100 170 580 Do 2 -60 -70 -70 -30 50 130 280 250 400 -40 250 CGO May 29 1 -70 -50 -30 130 190 300 370 360 390 140 230 540 Do 2 -80 -80 10 210 150 210 240 370 3C0 50 120 650 May 30 1 -70 -70 -70 -80 -50 -60 40 30 200 180 130 230 Do 9 -60 -60 -80 -40 -330 150 30 140 190 180 100 190 Col- . Date ony 5 6 7 8 9 10 11 12 1 2 3 4 num- p. m. p. m. p. m. p. m. p. m. p. m. p. m. p. m. a. m. a. m. a. m. a. m. ber May 18 1 470 210 -70 -50 -80 -60 -50 -50 -60 -50 -50 -70 Do 2 660 360 -50 -50 -70 -70 -70 -50 -60 -50 -50 -70 May 19 1 360 590 130 -100 -100 -90 -80 -100 -60 -70 -40 -80 Do. 2 470 710 280 -110 -100 -80 -80 -100 -60 -60 -40 -70 May 20 1 370 -130 -120 -70 -70 -60 -70 -90 -70 -60 -80 -70 Do 2 270 -130 -110 -90 -60 -80 -70 -70 -70 -70 -60 -80 May 21 1 290 350 -110 -120 -100 -90 -90 -110 -60 -80 -80 -70 Do 2 250 430 -130 -130 -120 -70 -110 -110 -70 -80 -80 -90 Mav 22. 1 490 240 -90 -80 -80 -70 -60 -60 -60 -60 -60 -50 Do 2 530 310 -100 -110 -70 -90 -80 -100 -60 -70 -70 -80 May 23.. 1 140 90 -60 -20 -50 -30 -40 -20 -40 -20 -30 -30 Do 2 180 160 -50 -40 -70 -30 -60 -30 -40 -40 -40 -40 May 24 1 460 370 -30 -70 -60 -70 -50 -50 -40 -50 -40 -50 Do.. 2 550 560 -30 -110 -60 -100 -40 -80 -60 -80 -60 -60 May 25. 1 400 400 70 -100 -90 -60 -70 -60 -50 -50 -50 -50 Do.. 2 580 560 -140 -130 -100 -80 -90 -90 -70 -70 -70 -80 May 26 1 400 480 360 -90 -70 -80 -50 -80 -50 -50 -50 -60 Do 2 550 640 440 -100 -90 -90 -80 -110 -60 -80 -80 -80 May 27 1 400 520 510 -100 -80 -90 -80 -80 -60 -60 -60 -60 Do 2 410 540 530 -90 -90 -100 -80 -120 -70 -60 -80 -70 May 28.. 1 560 750 490 -90 -100 -80 -70 -90 -60 -70 -70 -70 Do 2 610 760 450 -80 -100 -90 -60 -120 -70 -60 -80 -70 May 29 1 690 680 620 -80 -120 -80 -80 -110 -70 -70 -80 -80 Do 2 580 580 550 -80 -120 -60 -80 -110 -60 -60 -90 -80 May 30 1 520 710 150 -80 -80 -70 -50 -70 -50 -50 -50 -50 Do 2 440 620 140 -100 -50 -60 -50 -90 -30 -50 -60 ^50 The coefficient of correlation of the changes in weight between colony 1 and colony 2 for the diurnal hours from 6 a. m. to 7 p. m., inclusive, based on 277 hours, is .9070 ±.0071. The coefficient of correlation for the nocturnal hours, 8 p. m. to 5 a. m., inclusive, based on 190 hours, is .8938 ±.0098. The correlations are high, and have an insignificant probable error, indicating that the activity in both 14 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE colonies was chiefly determined by factors other than internal ones. The differences observed between these two colonies were almost ex- clusively the actual differences of loss or gain, and the most probable explanation of these differences is to be found in the presumably smaller number of bees in colony 1. Actual counts of the bees of the two colonies could not be taken without vitiating the experiment. If any structural difference existed between the bees of the two col- onies (as assumed by Merrill in his work), this was not determined, and with such high correlations as occur in the changes in weight of these two colonies such an explanation seems improbable. The high correlations do not indicate that any important internal difference, as in the condition of the queen, or the age of the bees, existed be- tween the colonies. If there had been such differences so high a correlation would not have existed. It must be kept in mind that .both colonies were amply supplied with storage and evaporating space. Fig. 5.— Graphs of the average variations, hour by hour, in the weight of Colonies 1, 2, and AB for the May honey flow. Note the minimum of the midday decline at 2 o’clock for the three colonies If one of the colonies had been crowded for space, and. this had inter- fered with the work of gathering, undoubtedly the coefficient of correlation would have been smaller. An examination of these data indicates that the field forces of the two colonies were working at approximately maximum efficiency. The two colonies gathered a total crop for the year 1923 which com- pares favorably with that of any other colonies kept in the general locality. An examination of the striking similarity of the changes in weight of the two colonies at the time when both regained their morning weight and at the time of the midday decline of increase in weight strongly suggests that both colonies were gathering all the nectar which °was available for them, in proportion to the number of bees available in each for field activity. This similarity is brought out much more clearly in the averages for the entire May honey flow (fig. 5) than in those for individual days (Table 1), since toward the WEATHER AND CHANGE IN WEIGHT' OF BEE COLONY 15 close of the honey flow colony 1 was apparently increasing in numbers more rapidly than colony 2. If it is assumed that both colonies were working at almost maximum efficiency, the midday decline in the rate of increase in weight to some extent agrees with Bonnier’s statements ( 1 ) that less nectar is available early in the afternoon than earlier or later. Unless there exists some unknown influence on bee behavior at the time of the midday decline, it must be believed that the reduction in the rate of the increase in weight at this time is due to a reduction in the amount of nectar in the honey plants. Surrounding the Bee Culture Laboratory there is a vast acreage of tuliptrees, and there are comparatively few bees in the neighborhood other than those of the bureau apiary. Bonnier’s assertion that bees carry partial loads of nectar when nectar is scarce is thus more prob- able, although the explanation may lie in the necessity for trips of longer duration at this time. There is no reason to believe that the bees actually gathered every drop of nectar available in the neigh- borhood at the time of this decline. THE SPRING PERIOD As has been stated, the data used in a consideration of the spring and fall honey-flow periods are not entirely comparable, since for the spring honey flow the only records used are those for days which show a net gain of at least 980 grams, whereas in the case of the fall honey flow the record for every day from September 4 to October 5 is used. Furthermore, the spring honey flow is much more intense in the vicinity of the laboratory than is that of the fall. For these reasons it seems best to consider the two periods separately. In the discussion immediately following, the various phases of the changes of weight during the day are taken up for the spring period, covering the time when the tuliptree was in bloom. MORNING LOSS Figure 3 illustrates the changes in weight during a typical day of the spring honey flow. In this case the morning loss is small and covers the time from 5 a. m. to 7 a. m., the majority of this loss occurring during the last hour. Usually the morning loss is quite small and rarely continues more than three or four hours. The amount and duration of the morning loss undoubtedly depend largely upon the proximity of nectar-producing plants and the abundance of their secretion, and upon weather factors prevailing at this time of the day. On some days the morning loss is negligible and scarcely distinguishable from the nocturnal loss, as on May 27 and 29, 1923 (fig. 6, 5). The bees began to return to the hive almost immediately on these days, and thus with their increased weight more than com- pensated for the loss during the early morning hours. It is quite evident that on such days nectar was abundant and within easy reach of the bees. The coefficient of correlation between morning loss and net gain is — .6350 ±.0652. This indicates that the smaller the morning loss the greater the resulting gain, and emphasizes the importance of locating an apiary as near as possible to the principal sources of nec- tar. The significance of morning loss will become more apparent later, in the consideration of the fall period. <3X3*?A-/^ 16 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE Fig. 6 (a).- 7VA*£r 0/=~ &s?V' s9M S> /O // /g' 7 Z 3 & -Daily graphs for six consecutive days of colony 2, showing hourly changes in weight (see Table 1) WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 17 Fio. 6 (6).- 77AOT 0/r£>^?y' 5 6 7 a 9 to U 12 7 2 3 4- 5 6 7 8 9 /O // /2' 7 2 3 4 5' +500 +400 + 300 +200 +/oo o -zoo + 700 +600 +500 +400 +300 +200 +/oo O -/OO -200 + 700 +600 + 500 +400 +300 +200 +700 O -/OO +600 +500 +400 +300 +200 +/oo O -/OO -Daily graphs for :even consecutive days of colony 2, following those oi hourly changes In weight (see Table 1) Kig.iru <> (a), showing 42201—25 ■3 18 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE MIDDAY DECLINE IN RATE OF GAIN During the period of spring honey flow, after the morning loss in weight has ceased, which may occur at any time between 6.30 and 10 a. m. (the average hour for 1922 was 8 a. m., and for 1923 9 a. m.), the colony increases in weight with a more or less regular accelera- tion until midday. At this time a decided slackening usually occurs in the rapidity of gain in weight. This change in rate of gain is sur- prisingly constant in the time of its appearance and can be distin- guished during practically the whole of the spring honey flow. The minimum rate of gain in weight, after gains have been established in the morning, usually comes at about 1 or 2 o’clock in the after- noon. On some days this decrease in rate of gain is decided, while on other days it is scarcely noticeable. When actual changes in weights are plotted this midday decrease in rapidity of gain can scarcely be seen, but when the data are plotted so as to show differ- ences in weight from hour to hour it becomes quite obvious (figs. 3,7,8). The cause of this change in rate of gain at midday is not entirely clear. Dufour (10) in his records mentions a similar phenomenon, and Bonnier (1, p. 163 ) connects a decrease in the amount of available nectar with low relative humidity and high temperatures prevailing at the same time of day. Bonnier measured the amount of nectar pro- duced at various hours and found that less nectar is produced toward the middle of the day, when high temperatures and low relative humid- ities are usually encountered. He further found that at this time bees return to the hive with less than their maximum loads, and that fewer bees leave the hive for the field during this midday period, in comparison with periods before and after. Either of these two facts might be sufficient to account for the midday decline in rate of gain. Figure 5, as stated, shows the average hourly changes in weight for colonies 1, 2, and AB for the spring honey flow. The similarity in the graphs for colonies 1 and 2 is not surprising, since these colonies were side by side and each was stimulated by exactly the same outside factors. The graph for colony AB, however, represents the changes for a colony one year earlier, but it closely follows the others. The midday decline in rate of gain is prominent in all three, and, although differing in magnitude, the three are intimately similar from hour to hour. The graphs for average temperature, relative humidity, and hours of sunshine bear no visible relation to the break in the three graphs showing gain and loss. The maximum temperatures for both years come later in the day than the midday decline, and the minimum relative humidity occurs at an hour slightly preceding. Until further investigations are made, Bonnier’s ( 1 , p. 164) observation of decreased nectar secretion on the part of the plant and the resulting effect on bee behavior must be accepted as the most logical explanation of this phenomenon, although the reasons which he gives for these changes on the part of the plants and the bees do not seem entirely satisfactory. NET GAIN The midday decline in rate of gain is included as part of the net gain, since an actual loss rarely occurs at this time during a good honey flow. The 13 graphs in Figure 6, for as many days, show WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 19 sharply how decidedly the midday decline reduces the amount of net gain. If the gaps caused by the midday decline were bridged the net gains would be appreciably larger. By theoretically bridging together the two peaks ol the graphs of hourly gain, and thus eliminating the midday decline in rate of gain, the net gain for colony AB is increased 15.72 per cent. The net gains of colonies 1 and 2 are similarly increased 16.35 per cent and 16.95 per cent, respectively, so that in Fio. 7. — Graphs of average hourly weight changes, temperature, relative humidity, and total hours of sunshine. Colony AB, spring period the years recorded the midday decline caused a loss of from 15 to 16 per cent in the net gain for the day. Actual gains in weight do not occur immediately after the sun rises, but they do cease almost as soon as the sun goes down, as shown in Figures 7 and 8. Although bees make actual gains in hours recorded as cloudy, it is interesting to note that in 1922 the actual hours of sunshine during the spring honey flow totaled 124.4, while 20 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE increases in weight were recorded during 136.4 hours. In May, 1923, during the honey flow there were 138.5 hours of sunshine, while colony 2 made gains during 137.5 hours. The similarity between the hours of sunshine and hours in which gains were made are somewhat of a coincidence, for, as stated, gains are sometimes made in cloudy hours and gains are not made in all hours of sunshine. The effect of sun- shine upon net gain will be discussed later (p. 36) . 7~//L*£T O/^ FIG. 8— Graphs of average hourly weight changes, temperature, relative humidity, and total hours of sunshine. Colony 2, spring period The temperature at the time when the colony ceases to gain in weight at the close of the day (figs. 7 and 8) is well above that at which bees are able to engage in effective outside work. This is shown by the fact that the average temperatures at the time of the morning loss, due to flights, are much lower than the temperatures WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 21 at the time that flights cease. The daily gains generally end abruptly, a fact suggesting that the bees do not desert the field at this time for lack of nectar. NOCTURNAL LOSS The nocturnal loss is calculated from the time that gain ceases until the bees leave for the field the following morning (fig. 2, D-E) . Nocturnal loss is largely the result of the evaporation of water which is given off in the process of ripening honey. In this regard, noctur- nal loss bears an intimate relation to the net gain. The coefficient of correlation between noctural loss immediately following the day’s gain and the net gain is .4112 ±.0909. The correlation of the diurnal gain with the loss of the second night following is much less, being .1S90±.1099. In the second instance the coefficient of correlation is not twice its probable error, and so can not be considered as especially significant. This would indicate that the greater part of the necessary evaporation is accomplished during the first night and that little is left over until the following night. Evaporation naturally begins almost as soon as nectar is fathered, and therefore occurs during the day of gathering and pro- ably during the following day, but in neither case is it possible to determine the amount of evaporation by day from a record of colony weights. In 1923, colony 1, from May 18 to May 30, lost during the nights 25.29 per cent of the total amount gained. Colony 2 in the same period lost 24.69 per cent of the total gain. For 11 days in May, 1922, colony AB lost during the nights 17.85 per cent of the total gain. This lower percentage of loss in 1922 may have been occasioned by the character of the nectar collected, since considerable locust honey was available. It is also well known that the water content of nectar varies from year to year with different varieties of plants and with climatic factors. A calculation of Dufour’s (11) data shows a loss of 26.16 per cent for 14 consecutive nights in May and June, when the minimum net gain was 970 grams, and the coefficient of correla- tion between net gain and nocturnal loss for this period is found to be .6663±.1002. Maujean (23) found a nocturnal loss of 22.53 per cent of the net gain during the honey flow of 1905, and a loss of 27.40 per cent during that of 1904. A calculation of Maujean’s data shows a coefficient of correlation of .7868 ±.0411 between the net gain and the following night’s loss, a correlation of .5319 ±.0795 between the daily net gain and the loss of the second night following the day con- cerned, and of .2813±.1035 between net gain and the loss of the third night following. These results would indicate that in this particular instance evaporation was practically complete by the end of the third night; a much slower evaporation than seems to have occured in the present investigation. Hommell (18) states that ac- cording to Sylviac seven days is the minimum observed time required to change nectar completely to the consistency of honey; he thinks, however, that seven days is entirely too long for this minimum period, and points out that de Layens long ago discovered that in warm and dry weather nectar may be of such consistency as to permit almost immediate capping by the bees. Hommell points out that Huillon likewise gives a much shorter period for the completion of evapora- tion, and states that beekeepers can remove honey during the morn- 22 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE ing following a heavy honey flow without much danger that the honey will be too thin, since the greater part of the evaporation takes place during the first night. The strength of the colony must likewise be an important factor in the rate of evaporation. While it is true that a weak colony collects less nectar than a strong one, it does not follow that the efficiency of the two colonies is proportional to their strength. In studying the correlations in Table 2 it is seen that the external factors have but little influence on nocturnal loss. If there existed a strong internal individual influence in either one of these colonies, one would not expect such a high correlation between the nocturnal losses of colonies 1 and 2 (.8938, see p. 13). If external factors are important so far as nocturnal loss is concerned they must be other than temperature, temperature variation, relative humidity, or variation of relative humidity. Again the need for strong colonies is apparent. Weather conditions at night undoubtedly influence weak colonies to a greater extent than they do strong ones. Phillips and Demuth (30) in their wintering experiments have shown the detrimental effect of low temperatures upon weakened colonies, while strong colonies reacted normally to the same temperatures, if we take the survival of strong colonies and the death of weak ones as a criterion for differences in normal behavior. THE FALL PERIOD The data upon which the calculations for this period are based were secured from September 4 to October 5, 1922, during the period of the fall honey flow. No criterion for minimum gain was used in selecting the day to be correlated, and all days irrespective of gain or loss are included in the calculations. For this reason the figures can not be compared directly with those of the spring period, when only days having a minimum of 980 grams were selected. In studying the figures representing days of the fall period, it will be seen that the graphic presentations of hive weights are more signifi- cant than graphs showing the difference in weight from hour to hour. As in the analysis of the hourly weight changes of the spring period, weather factors, as they affect changes of weight in the fall, will be postponed for later discussion. MORNING LOSS The relation of morning loss to net gain appears to be the reverse of what it is during the spring period. The correlation between morning loss and net gain is .5769 ±.0825, so that not only is the correlation between these two factors numerically greater than it is in the spring, but it is positive rather than negative. The morning losses generally were heavy, being from 390 to 1,690 grams. The morning loss is primarily brought about by the rapid exit of bees, as undoubtedly the returning bees with their loads of nectar more than make up for the loss due to the consumption of stores for maintenance of the colony, which constitutes a steady loss at all times. The duration of these losses varied from four to nine hours, and often equally as long a time would be required for the colony to regain its original weight of the early morning, this in turn leaving only WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 23 77^£~ o/=~ Fig. 9.— Ilourly changes in weight of colony B; A, September 23, B, September 25, 1922 24 BULLETIN 1339, U. S. DEPARTMENT OP AGRICULTURE a '"few hours at the close of the day for the accumulation of net gain. 'J^On September 25, 1922, colony B sustained the maximum morn- ing loss of the season (fig. 9). The energy expended by the bees on this day, from the beekeeper’s point of view, was misappropriated since the result of the day’s labor netted the colony a loss of 255 grams. The weather of this day was ideal in every respect, and the bees evidently searched for nectar but found little or none. The average morning loss of the spring period is insignificant in com- parison with that of the fall period. Figure 9 illustrates the mag- nitude of the morning losses on September 23 and 25, 1922, respectively, while Figure 10 shows the average loss for the entire fall honey flow of that year. 77/W£: Fig 10. -Graphs of average hourly changes in weight, temperature, relative humidity, and total hours of sunshine. Colony B, fall period MIDDAY DECLINE IN RATE OF GAIN Although it is possible to distinguish the midday decline in rate of gain in weight during the majority of the fall days, it is seldom pronounced in its character. It is shown in the shaded portion of Figure 10, which presents the average variations in weight for each hour of the day in the fall period. The graphs of actual weights of fall days do not show this midday decline because it occurs so shortly after the time when the colony stops losing weight. The effect of the midday decline, as determined by bridging the gap at this period upon the net gain, is, however, greater than it is during the spring, since in this case the net gain was decreased 50.66 per cent on account of it. WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 25 NET GAIN The net gain is accumulated late in the day. Figure 10 shows that the colony did not regain its original weight until between 4 and 5 o’clock in the afternoon, and since increase stops on the aver- age at 6 o’clock any net gain must he accumulated between these hours. Undoubtedly the nature of the nectariferous plants abound- ing in any one locality will, together with the prevailing climatic factors, determine the time and amount of gain in weight. An en- tirely different picture may be expected in the buckwheat regions, since buckwheat is known to secrete nectar only in the forenoon. Figure 1 well illustrates the vast differences, as shown by hourly changes in the hive weight, that may exist in the interrelationship of nectar-producing plants, bees, and weather factors. The two curves in this figure represent these changes for May 9 (solid line) and September 28 (dash line), 1922. On both of these days the net gain was practically identical, but the activities of the bees in secur- ing this gain were vastly different on the two occasions. On the earlier day the morning loss ceased at 8 o’clock; on the other it contin- ued until 11 o’clock and was more than seven times as great. On May 9 the original weight was regained by 12 o’clock, whereas on September 28 it was not regained until between 4 and 5 o’clock in the afternoon. The comparison of these two days also serves to em- phasize the statement made earlier that to be of service weight rec- ords must be taken frequently and methodically. If only early morning and evening records are taken nothing can be known as to the time during the day when nectar is being brought into the hive, and, moreover, if the last weight of the day were made an hour or two earlier, one day might show a gain while the other shows a loss. NOCTURNAL LOSS The nocturnal loss during the fall is appreciably higher in com- parison with the net gain than it is in the spring. For the days showing a net gain in Table 3, the nocturnal loss amounts to 48.72 per cent of the net gain. The correlation between these two is also much higher than it is in the spring, being .8568 ±.0333. A calcu- lation of Dufour’s (11) data likewise shows a higher correlation be- tween nocturnal loss and net gain for the fall, being .8689 ±.0458, as opposed to a spring correlation of .6663 ±.1002. A higher correlation between net gain and nocturnal loss, as well as a larger per cent of loss, is naturally to be expected when net gains are small. Such a condition obtains in this case, since a larger part of the nocturnal loss is to be attributed to consumption for colony maintenance. CORRELATIONS BETWEEN EXTERNAL FACTORS AND THE CHANGES IN COLONY WEIGHT Having discussed the changes in weight of the colonies during the spring and fall honey flows and the correlations for the various parts of the day, it now is desirable to determine what degree of correlation exists between the changes in weight and the factors of the environ- ment in so far as data are available for computation. These will be discussed in the following order: Temperature, relative humidity, solar radiation, and hours of sunshine. Numerous coefficients of correlation along these lines are presented in Table 2, together with their probable errors'. 42201—25 1 Table 2. — C oefficients of correlation between changes of colony weight in the spring and various weather factors 26 BULLETIN 1339. U. S. DEPARTMENT OF AGRICULTURE Noc- turnal relative humidity variation 0. 0703 ±. 1087 -.0654 ±. 1089 -.0157 ±. 1093 CO CO tO O <“I CO lO o 41 -.7018 ±. 0555 .5294 ±.0553 . 6343 ±. 0653 .6258 ±. 0665 -.3373 ±.0969 .7031 ±. 0553 -.6819 ±.0585 Noc- turnal average relative humidity -0.4495 ±.0873 1264 ±. 1076 CM © OO .M OO CD CO o 1 41 -.5114 ±. 0808 .5833 ±.0721 -.5344 ±.0781 -. 6445 ±. 0039 00 to O CO CO to o 1 -H -. 0043 ±. 1094 -. 8781 ±.0250 -.6819 ±.0585 Noc- turnal tempera- ture variation 0. 2941 ±. 0999 cd CO CD CD co o 1 -H .2571 ±. 1021 CD o ■'T* CD CO o -H -. 4264 ±. 0895 . 1970 ±. 1051 .3822 ±. 0934 00 o to u* CO CD CO o 41 -. 0739 ±. 1088 -.8781 ±.0250 .7031 ±.0553 Noc- turnal average tempera- ture 0. 5510 ±. 0702 r* © O 41 .7947 ±. 0403 .3185 ±. 0983 . 1956 ±. 1052 -. 0072 ±. 1094 CD L— <-D o p 41 . 1012 ±. 1082 -.0739 ±. 1088 -. 0043 ±. 1094 -. 3373 ±. 0969 Solar radiation 0. 5525 ±.0760 .2375 ±. 1032 .4416 ±. 0880 L- CD CM CD o o 41 -.6518 ±. 0629 .6280 ±. 0662 CM CO CM 00 o -H . 1012 ±. 1082 .3358 ±. 0970 -. 5308 ±.0785 . 6258 ±. 0605 Hours sunshine 0. 6124 ±.0083 .3068 ±. 0991 .5798 ±. 0726 .7703 ±. 0444 -. 6420 ±. 0643 . 6780 ±.0591 .8826 ±. 0241 .2059 ±. 1047 .3822 ±.0934 -. 6445 ±. 0639 .6343 ±. 0653 Diurnal relative humidity variation 0. 4229 ±. 0898 . 1053 ±. 1004 .3899 ±. 0927 .7852 ±.0419 CO CD £> O 1 -H .6780 ±.0591 .6280 1 ±.0602 | CM rf U- 03 o o O T-H 1 41 O — t>. to CD O 41 -. 5344 ±. 0781 .5294 ±. 0553 Diurnal average relative humidity -0. 3806 ±. 0935 . 0141 ±. 1093 -. 1551 ±. 1007 -.7272 ±.0515 -.7584 ±. 0464 -. 6420 ±.0643 CO CD ^ CM tO CO O O 1 4i . 1956 ±.1052 © © CM CO O 1 41 . 5833 ±.0721 -. 7018 ±. 0555 Diurnal tempera- ture variation 0. 5907 ±. 0704 .0583 ±. 1090 t>. zo — I CO u- u- o o 41 CM O L- ^ CM O O 1 41 CM 03 CO 'rT r- o -H .7703 +. 0444 D- CD CM C3 P- tO CD © 41 . 31S5 ±.0983 .3419 ±.0960 -.'5114 ±. 0808 .5158 ±. 0803 Diurnal average tempera- ture 0. 7529 ±. 0473 .3063 ±. 0991 t- O *-i CO T- r- to o 41 -. 1551 ±. 1067 .3899 ±. 0927 .5798 ±. 0726 CD O r-H CO -r-i CO ^ O 41 .7947 ±. 0403 .2571 ±. 1021 -. 3882 ±. 0929 -. 0157 ±. 1093 Noc- turnal loss 0.4112 ±. 0909 .3063 ±. 0991 .0583 ±. 1090 .0141 ±. 1093 . 1653 ±. 1064 .3068 ±. 0991 tO CM p- :o CO o CM —i 41 . 1754 ±. 1060 -.3439 ±.0964 -. 1264 ±. 1076 -. 0654 ±. 1089 Net gain 0. 4112 ±. 0909 .7529 ±.0473 .5967 ±. 0704 O to O CO 00 05 CO o 1 41 .4229 ±. 0898 .6124 ±. 0683 «o o CM O tO U- to o 41 . 5510 ±. 0762 .2941 ±.0999 -.4495 ±. 0873 .0763 ±. 1087 .5 'c3 txO © m Nocturnal loss Diurnal average temperature Diurnal variation of temperature Diurnal average relative humidity Diurnal variation of relative humidity Hours sunshine Solar radiation Nocturnal average temperature Nocturnal variation of temperature Nocturnal average relative humidity Nocturnal variation of relative humidity WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 27 TEMPERATURE AND CHANGES IN THE COLONY WEIGHT In the vast literature of beekeeping few data are found to be useful in throwing light on the relation of external temperature to the changes in colony weight. Many experiments, however, have been conducted to ascertain the relation of temperature to nectar secretion. Of the latter, the work of Bonnier stands out most prominently. Unfortunately, he published but few of the actual data upon which his conclusions were based, so that it is difficult to make mathemati- cal comparisons with the data herein presented or with the data of other investigators. Broadly speaking, the correlation between plant or animal activity and temperature is positive within, of course, cer- tain limits. Most reactions, whether organic or inorganic in nature, are retarded by low temperatures, and increased by higher tempera- tures, although there are many examples to the contrary. TEMPERATURE AND NET GAIN During the spring period temperature has a marked effect upon net gain, the correlation between these factors being .7529 ±.0473. This is a higher correlation than was found between net gain and any of the other weather factors, from which it appears that among those considered, temperature is the most important single factor influenc- ing changes in the colony weight. Bonnier ( 1 , p. 163), in referring to nectar secretion, states that the volume of nectar varies inversely as the temperature. Kenoyer {20) also says that the accumulation of sugar in the flower and its vicinity varies inversely as the temper- ature. In referring to honey production, however, Kenoyer {19) states that good honey months average slightly higher in temperature than poor ones, this being especially true of the fall and spring months, and that the yield is best on days having a maximum of 80° to 90° F. (.26.7° to 32.2° C.). Ono {26, p. 15) , in studying the secretion of extra- floral nectaries, does not lay much stress on temperature. He found that within the range of 15° to 25° C. (59° to 77° F.) the influence of temperature upon secretion is not remarkable; and that tempera- ture seems to have merely an indirect influence in so far as it affects the plant itself. Harrault {15) states that for a given plant placed in suitable vegetative conditions the production of nectar increases ■with the increase in temperature and the quantity of sunlight, the production following general botanical laws. The writer’s calculation of ITarrault’s rather scanty data gives a correlation of .5258 ±.0639 between net gain and temperature. A negligible correlation of .0065 ±.1079 was likewise calculated from data presented by Mau- jean (23). Bonnier reached his conclusion by removing the nectar from vari- ous species of plants from hour to hour by means of a small pipette, and noting the temperature. The writer’s calculations of Bonnier’s data gave a correlation of — .6581 ± .0901 between temperature and nectar secretion . Coefficients of correlations were also calculated from the same data between humidity and gain and between humidity and temperature, with the following results, where temperature, H— humidity, and N= volume of nectar: rNT= - .6581 ±.0901 rN[I= .8040 ±.0562 rT,i= - .9202 ±.0243 28 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE Humidity is, of course, influenced by temperature. In order, there- fore, to arrive at a more correct value of the effect of temperature upon nectar secretion, it is necessary to eliminate the effect of humid- ity by reducing this variable to a constant. This is conveniently done by the employment of the formula for partial correlation — T NT — {TnH • T Til) uTnt ~ va-rwHi^*) Substituting in this formula the above coefficients of correlation, and solving, we find that — hTnt = .3306 It seems, therefore, according to Bonnier’s data, that when humidity is constant nectar secretion varies directly with temperature. The interpretation of similar data compiled by other investigators would be highly desirable at this place, so that results could be com- pared with those of Bonnier. The writer’s calculation of Harrault’s (15) data gives the following results: Tho = .2988 ± . 1 228 rrG=. 5258 ±.0639 rnT = . 2821 ±.1241 In this case the net gain of a colony of bees was used, rather than the volume of nectar, and G therefore equals net gain. By substi- tuting these values in Pearson’s formula and solving we find that — - hTtg — .4826 The coefficient of correlation between net gain and temperature after correction is, therefore, somewhat lower than the direct correlation, but it is still positive. In Table 2 the following coefficients of correlation are found: Thg = — .3806 ± .0935 rTG= .7529 ±.0473 rTH =* - .1551 ±.1067 In this case, by reducing the effect of humidity to a constant, we have — hV to = .7594 No attempt has been made in the present investigation to ascer- tain the optimum temperature for the maximum change in weight of a colony of bees. Kenoyer (20) found under experimental conditions that the optimum temperature for nectar secretion for most of the Leguminosae which he tested was 15° C. (59° F.). He also called attention to the fact that the phenomenon of the accumulation of stored sugar from starch at low temperature in the twigs of woody plants is well known, and, further, presented data which show that the same is applicable to floral tissues. He likewise ascertained that, WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 29 with an increase in temperature, the permeability of the protoplasts ■of the multicellular secreting hairs which cover the nectary of Abu- tilon rapidly increases. The osmotic pressures (intimately related to permeability in plants) of aqueous •sucrose solutions were like- wise found by Morse, as cited by Stiles (32) , to vary directly with temperature, at temperatures generally encountered in working with plants. With the effect of low temperatures upon the accumulation of sugar on the one hand, and the effect of high temperatures upon permeability on the other, Kenoyer states that the evidence points to the conclusion that the secretion of nectar results from a balance between the two, and that the optimum temperature for secretion may be represented by the point where the negative graph repre- senting temperature and sugar accumulation crosses the positive graph of permeability of protoplasts to sugar and temperature. Phillips and Demuth (31), in referring to white clover (Trifolium repens), state that this species of Leguminosae may rarely be counted upon as a major honey source where the average summer tempera- ture exceeds 75° F. (23.9° C.), and a more important consideration is that secretion is most rapid where there is a considerable daily range of temperature, the best results being observed when the night temperature is below 65° F. (18.3° C.), and the day temperature above that. The slight difference between Kenoyer’s optimum tem- perature and the observations of Phillips and Demuth may be •explained by the fact that the former was studying nectar secretion •while the iatter were referring to conditions as measured by the honey crop. There are numerous records where low temperatures appear to have an important bearing upon nectar secretion. On the other hand, the literature of beekeeping contains many references, mainly •observations by beekeepers, to the effect that high temperatures are necessary for an abundant secretion of nectar. Undoubtedly the physiological behavior of plants in nectar secretion varies somewhat with different species and under varying climatic conditions. Like Ono (26) , Wilson (33) found that temperature makes but little dif- ference in nectar production. In the case of branches of Prunus laurocerasus , however, he discovered that a temperature of at least 12° C. (53.6° F.) is necessary for the metamorphosis of the cell walls and the raising of the cuticle, and that after this activity of the nec- tary had passed a much lower temperature sufficed for continued secretion. Haupt (17) found that a certain minimum temperature was necessary to induce secretion and Demuth (9), in making obser- vations to determine the temperature at which basswood (Tilia americana) begins to yield nectar, found that in northern Indiana this plant did not yield nectar until a temperature of 64° F. (17.8° O.) was reached. From what has preceded, it is quite evident that considerable con- fusion exists regarding the true relation of temperature either to nectar secretion or to changes in colony weight. Much of this con- fusion is undoubtedly due to a lack of sufficient discrimination be- tween the uses of the terms “nectar secretion” and “change in colony weight.” So far as the changes in weight aro concerned, the majority of the data indicate that during a honey flow the relation of these changes to temperature is positive. If there does exist a negative correlation between nectar secretion and temperature, it 30 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE appears evident that either the honey flora in the vicinity of Somer- set, Md., does not respond during the spring period as did the plants studied by Bonnier and Kenoyer, or else that the temperature has such a pronounced effect upon the behavior of the bees that the effect upon nectar secretion is largely obscured. Hot, dry weather often occurs at periods when no nectar is being brought to the hive; at the hottest times, however, it is not at all uncommon for bees to gather large quantities of honeydew of insect origin. Since the bee is a cold-blooded animal it is but natural to expect increased activity with increasing external temperature. Table 3 lists the coefficients of correlation between net gain and the various weather factors of the previous day. It will be seen that the coefficient of correlation between the net gain and the average temperature for the entire spring period is .1783 ±.1059. This in- fluence is positive but not otherwise especially significant, since the probable error is high. The correlation of net gain with average nocturnal temperature just preceding the gain is slightly higher, being .2151 ±.1043. Evidently, according to the data presented, tempera- tures of the day on which gains are made have a more direct influence on colony gains than do those of the previous day or night. Table 3. — C oefficienls of correlation betiveen net gain in colony weight and various weather factors of the 'preceding day, with their mean and standard deviations for the spring period Diurnal average temper- ature Diurnal varia- tion in temper- ature Diurnal av erage relative humid- ity Diurnal varia- tion in relative humid- ity Hours sun- shine Solar radia- tion Noc- turnal average temper- ature Noc- turnal varia- tion in temper- ature Noc- turnal average relative humid- ity Noc- turnal varia- tion in relative humid- ity Net gain 0. 1783 0. 3317 0. 0329 0. 0289 0. 0246 -0. 0804 0. 2151 0. 2472 0. 0005 0. 0721 ±. 1059 ±. 0973 ±. 1093 ±. 1093 ±. 1093 ±. 1087 ±. 1043 ±. 1027 ±. 1094 ±. 1088 Mean deviation Standard deviation.. 4. 5261 2. 4895 4. 9735 2.3113 3. 1578 2. 3230 4. 6577 2. 4526 5. 8682 2. 7067 6. 4998 2. 4146 5. 3682 2. 8603 4. 6577 1. 8416 5. 6840 2. 0148 4. 1314 2. 3304 The data taken during the fall period give a correlation coefficient of —.2310 ±.1185 between the average diurnal temperature and the change in weight. This figure is radically different from the corre- lation of .7529 existing between the average diurnal temperature and the net gain for the spring period. The probable error of the fall correlation is large, and this figure is therefore important chiefly because there is a negative correlation. This difference may be due in part to the fact that during the fall the activity of the colony in brood production is carried on under somewhat adverse conditions; that is, the effort necessary properly to care for the brood during the fall is proportionally greater than it is during the spring, and for this reason the bees will go to the fields under conditions otherwise not ideal. In such a case it may be that temperature has a more pro- nounced effect upon nectar secretion than upon bee behavior, and thus results in a negative correlation between average diurnal temper- ature and changes in colony weight. There seems little doubt but that a much greater effort is necessary on the part of the bees during the fall properly to carry on normal colony activity than is needed during the spring. Figure 1 illustrates a day in each of these two WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 31 periods when the net gains are approximately identical. For the spring day, the morning loss was small and ceased at 8 o’clock. From this time on the colony gained slowly but steadily until the close of day. Apparently only partial loads were carried during most of the day, and the bees made short trips to the field. Had either one of these two suppositions not been true, the morning loss would have been larger, or, if the bees had carried maximum loads and had made short trips, only a small portion of the field force would have been engaged in gathering. The fall day shows a tremendous morning loss with a regaining of the original weight late in the afternoon. Such a heavy loss, scattered over so many hours, tvnuld justify one in believing that the entire field force was active, and that the indi- vidual bees made long trips and perhaps carried but partial loads. Figure 9 ( B) illustrates another day (September 25) of the fall period when the bees must have made tremendous efforts to secure what nectar was available. Table 4 presents various data relating to the fall honey flow of 1922. Table 4. — - Changes in weight of colony B, and various weather data during the fall honey flow Date (19221 Diur- nal change in weight Noc- turnal loss Diur- nal aver- age tem- pera- ture Diur- nal tem- pera- ture varia- tion Diur- nal aver- age rel- ative hu- midity Diur- nal rela- tive hu- midity varia- tion Sun- shine Solar radia- tion Noc- turnal aver- age tem- pera- ture Noc- turnal tem- pera- ture varia- tion Noc- turnal aver- age rel- ative hu- midity Noc- turnal rela- tive hu- midity varia- tion Morn- ing loss Sept. 4 Grams -130 Grams 80 °F. 87.5 °F. 20 69.3 44 Hours 11.7 438.3 °F. 77.5 °F. 9 97. 1 5 Grams 450 5 -70 60 85.3 18 68.6 53 9.6 432.6 73.2 12 97.7 8 420 6 -50 60 87.0 26 62.2 60 12.8 490.6 74.9 13 95.3 11 460 7 -10 50 87.6 24 64.5 56 8.7 450. 1 78.0 12 96.0 9 390 9 -110 90 81.4 13 79.8 39 5.2 277.5 74.7 7 98.8 4 690 10 200 130 83.9 15 68.6 47 8. 1 410.9 75.7 6 97.7 10 410 12 -20 20 75. 1 9 71.6 54 6.3 301.4 61.0 14 97.5 10 820 13 290 120 74.6 27 54.5 64 12.5 495.2 64.9 9 96. 1 9 800 14 550 200 81.9 27 58.0 48 12.5 478.9 74.0 11 95.5 14 850 15 550 170 84.7 21 66.3 55 11. 1 464.4 74.8 9 97.5 6 920 17 50 90 69.6 19 56.0 55 7.9 308.0 55.7 11 95.6 8 1,010 18 90 170 69.0 22 53.3 36 10.4 413.6 61.9 5 90.8 9 1,080 19 370 190 70.5 18 63.3 51 5.7 271.9 61. 1 13 97.7 13 1,170 20 380 190 66.9 18 76. 1 45 6.4 289. 1 62.5 8 98. 1 8 1,185 21 690 290 70.7 15 68.6 48 3.7 311.9 61.5 13 97.0 11 1, 310 22 410 235 69.6 22 68.5 57 3.2 323.2 60.5 14 98.3 6 1,510 23 1,395 390 77.4 33 58.2 72 10.4 397.3 65.8 17 96.9 10 1,430 24 960 510 78.3 32 66.5 62 6.0 329.8 68.4 23 71.5 49 1,620 25 -220 170 67.7 18 46.2 71 12.0 540. 0 50. 1 13 91. 1 19 1,690 26 310 220 61.2 25 49.2 70 12.0 474.5 51.2 10 96.7 20 1,360 27 1,110 340 68.9 32 60.0 60 12.0 448.9 61.6 13 96.5 17- 1,260 1,495 28 680 300 71.3 34 58.8 70 11.7 404.3 61.2 10 97.0 15 29 300 230 72.9 22 71.9 51 7.9 346. 5 62.0 17 97.8 7 1,300 30 660 300 73.0 28 64.5 56 11.8 413.5 60.2 16 96.9 11 1,190 Oct. 1 440 230 74.4 30 61.0 64 11.8 431.5 58.8 6 97. 1 11 1,190 1, 150 2 380 220 75.2 34 63.4 66 11.6 396.8 60. 1 16 98.5 7 3 350 215 76.6 34 62.0 64 11.7 361. 1 65.5 13 97.7 8 1,070 4 215 210 78.0 27 63.9 60 10.8 306. 6 69.5 12 95.5 9 1,005 5 240 225 83.0 32 53. 5 74 11.6 365.5 69.6 14 87.0 23 900 It must be recalled that the data in the spring and fall periods are not entirely comparable, since days were not chosen in the fall with a minimum net gain of 980 grams. The difference between the correlations of temperature and net gain in the two periods is, how- ever, so striking as to call for an explanation, and no definite ex- planation seems to be available. In determining the cause of this difference in correlation in the two periods it must be remembered 32 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE that there are striking differences in the conditions of the colony; in the spring the brood rearing is increasing and the population of the colony is not only greater but is made up predominantly of young bees; in the fall the population is smaller, brood rearing is decreas- ing, and many of the field bees are older than in spring. In the spring in the vicinity of the Bee Culture Laboratory the honey flow is heavy and nectar is abundant, whereas in the fall the daily gain is usually small and the plants yielding nectar do so in relatively small quan- tities. There may be an undetermined difference in the time of day at which nectar is secreted (as is found between clover and buck- wheat, for example) . In the spring, when nectar is unusually abun- dant, the gains may be largely attributable to the influence of increased temperature on bee activity, whereas in the fall the greater gains at lower temperatures may be associated with the inverse re- lationship which has been assumed between nectar secretion and temperature. Until more observations are made this must remain a matter of speculation. It will be noticed in Table 2 that the correlation between diurnal variation of temperature and net gain in spring is .5967 ±.0704. A wide daily variation in temperature on days of good gathering has long been noted by beekeepers. There is some evidence, as previously pointed out, that a preceding low temperature is necessary to induce nectar secretion, and that once started this secretion will continue at higher temperatures. The correlation of this same variable with the fall changes of weight (Table 5) gives approximately the same figure, namely, .5570 ± .0863. The fact that diurnal variation of temperature affects the weight changes alike, both in the spring and fall, and the observation of this phenomenon by beekeepers, would indicate that this factor has a greater influence on nectar secretion than it does upon bee behavior. For the fall period there is a correlation of — .7586 ± .0531 between the average diurnal temperature and morning loss. The morning loss is largely brought about by flight activity, and is therefore really a quantitative measurement of it. Since there exists for this period a positive correlation between morning loss and change of colony weight, this fact would seem to indicate that the colder the weather the greater the gain, as in reaiity the correlation between the changes of colony weight and diurnal average temperature (—.2310) actu- ally shows, though to a small degree. The correlation between diur- nal variation of temperature and change of colony weight is positive, and the correlation between morning loss and diurnal variation of temperature is also positive, despite the fact that there is a large negative correlation between morning loss and the diurnal average temperature. In calculating this latter coefficient of correlation, the average temperature for the entire diurnal period was used, while the period of morning loss covers but half of this time. Since the maximum average temperature comes later than the end of the per- iod of average morning loss, as shown in Figure 10, a correlation with the temperature for the hours actually coinciding with morning loss would undoubtedly result in a lower correlation between these two variables. The evidence in favor of the theory that during the fall temperature has a greater effect on nectar secretion than it has on bee behavior is thus not materially weakened. WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 33 TEMPERATURE AND MIDDAY DECLINE As previously pointed out, the cause for the midday decline has been attributed to high temperature and low relative humidity, both prevailing at that time of the day. The minimum of the average midday decline occurs from 1 to 2 p. m. (figs. 7, 8, and 10). The maximum of the average temperature comes from 3 to 4 p. m. (figs. 8 and 10), although it occurred an hour earlier during the 1922 spring honey flow (fig. 7) . Although the midday decline occurs during the hot part of the day, it does not coincide with the period of maximum temperature. Moreover, the coefficient of correlation between net gain and average temperature is high and positive for the spring period, so that it is difficult to give a satisfactory explanation of the occurrence of the midday decline on the basis of high temperature and low relative humidity. During the fall the relation between change of colony weight and the average diurnal temperature is neg- ative and the coefficient is small. From the standpoint of the net gain, the midday decline during the fall is large, but from the stand- point of weight changes it is small and barely noticeable. TEMPERATURE AND NOCTURNAL LOSS The relation of temperature to nocturnal loss is surprising in view of the widespread importance usually attributed to it by beekeepers. One naturally expects to find a higher rate of evaporation on warm nights following days of good gain than on cool nights. This relation, however, is small, the coefficient of correlation between net gain and nocturnal average temperature being but .1754 ±.1060. Temperature variation seems to have a more important bearing, the coefficient of correlation between net gain and nocturnal vari- ation of temperature being — .3439 ±.0964. One interpretation for this would be that when there is a great variation in temperature during the night the bees are forced constantly to modify their tem- perature-regulating organization in such a manner that the temper- ature of the brood-chamber may be maintained constant. In doing this the task of evaporation is interrupted, resulting in a negative correlation. In other words, from the standpoint of evaporation, an even temperature is more desirable than a high temperature, within, of course, reasonable limits. During the fall period the relation between temperature and noc- turnal loss is the opposite of that existing in the spring period, the coefficient of correlation between the average nocturnal temperature and nocturnal loss being — .3391 ± .1108. It would be contrary to physical laws to expect a higher rate of evaporation during cold nights than during warm nights. An explanation of this negative correlation seems to lie in the fact that the relation between changes of weight during the day and the respective nocturnal losses is high. This in turn would indicate that the incoming nectar was promp'ly cared for, and the existing outside temperature, instead of primarily affecting the rate of evaporation, has its effect principally upon the general activity of the colony, resulting in the consumption of larger quantities of stores on cold nights. 34 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE The coefficient of .5136 ±.0922 between nocturnal loss during the fall and nocturnal variation of temperature can be explained in the same way. During the fall the amount of evaporation necessary is comparatively small because of the small amount of nectar gathered during this investigation, so that the various weather factors have little effect upon it, but have a considerable effect upon the work of the colony as a whole. The greater the range in temperature, the greater the consumption of stores, owing again, perhaps, to modifi- cation in the temperature-regulating arrangements of the colony. It is seen that variation of temperature has a greater influence upon the nocturnal loss, during both the spring and fall, than does the average nocturnal temperature. Variation of temperature at night produces noticeable disturbance. RELATIVE HUMIDITY AND CHANGES IN COLONY WEIGHT Even less is known regarding the influence of relative humidity upon changes in the weight of the colony than regarding the influ- ence of temperature, probably largely because instruments for indi- cating relative humidity are not as widely distributed as are ther- mometers. Plant physiologists have, of course, studied the influence of relative humidity upon various phases of plant life, and to a lim- ited extent upon nectar secretion. As in the case of temperature, there is no general agreement as to its relative value. One of the difficulties encountered in determining the value of the influence of relative humidity upon any phase of plant or animal life is that relative humidity and temperature are intimately related. It is necessary, in a careful analysis of the influence of various weather factors upon physiological changes, to make due allowance for the influence of these variables upon each other. Wright (34), for instance, in recalculating the figures of Briggs and Shantz (4, 5) found that temperature and not. wet-bulb depression was the more important variable, influencing the daily transpiration of plants. Patterson (28) states that relative humidity as a factor influencing the growth of higher plants has been greatly overestimated. He found that elongation in etiolated shoots of the common bean, growing in pure silica sand, either 20 per cent or 60 per cent saturated with water, proceeded as rapidly in a relative humidity of 30 per cent as in a relative humidity of either 60 per cent or 90 per cent. When growing in silica sand, 5 per cent saturated, they grew less rapidly in a relative humidity of 30 per cent than in a relative humidity of 60 per cent or 90 per cent, but grew as rapidly in a relative humidity of 60 per cent as they did in one of 90 per cent. All figures pertaining to relative humidity in the present bulletin were obtained from a self-recording hygrometer maintained in the laboratory apiary. This instrument tended regularly to record too high percentages of relative humidity, and consequently the figures in some of the tables appear to be above the average. This error in the instrument should not greatly affect the coefficients of correlation in which relative humidity figured as one of the variables, since the error appears to be fairly constant. No attempt has been made to determine the optimum or minimum relative humidity necessary for any activity of the apiary. WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 35 RELATIVE HUMIDITY AND NET GAIN It has been shown that temperature influences nectar secretion and changes in the colony weight. Similarly, plants doubtless respond to changes in relative humidity according to the species and the location in which they grow. Some species of plants are known to se- crete nectar under conditions of the greatest possible humidity; on the other hand, arid and semiarid countries contain many nectar- producing plants. Hommell (18, p. 191^) states that, other conditions being equal, the quantity of nectar increases with an increase in the relative humidity ■of the air. Ono (26) , in discussing the influence of moisture on nec- tar secretion in extrafloral nectaries, says that “moisture is one of the conditions most favorable to the secretion of nectar, and its influence seems to be more or less direct. Dry atmosphere is in all cases un- favorable to the secretion/’ He cites a case in which cut twigs of Prunus laurocerasus were placed in a water bottle and held under a bell jar of moist air for three weeks. Despite the fact that these nectaries were washed daily (an operation which will often terminate secretion), they continued throughout the experiment to secrete nectar without decreasing either its quantity or the quantity of sugar contained in it. Wilson (33), working with the same species of plant, found that after a branch of Prunus laurocerasus had been made to secrete nectar by being placed under a moist bell jar, it continued to do so without water and in the dry air of an ordinary room, until the whole branch had lost more than one-fourth of its weight by withering. Kenoyer (20) discovered that by increasing the relative humidity the production of water was increased, but not that of sugar from nectaries. Bonnier (1 ) found that the secretion of nectar varies directly with relative humidity. Although he realized (2) the complex role which relative humidity plays in plant physiology, he made no corrections for the effect of other variables upon relative humidity in his studies on nectar secretion. In calling attention to the midday decline, and accounting for it by the low relative humidity and the high temper- ature prevailing at that time, he made no effort to assign more im- portance to one than to the other. The writer, in calculating Bonnier’s data, found the coefficients of correlation, as stated under the discus- sion of temperature. In order to arrive at a corrected value for the effect of relative humidity by reducing temperature to a constant, we substitute in the following formula the numerical values of the proper coefficients (see p. 27 for correlation values), and obtain — Tbh — (rNT- rm) TrNH = -yJ(l-r2NT)(l:zr\B) = .6605 Thus the effect of temperature reduces the coefficient of .8040 between relative humidity ancl nectar secretion to .6605. The substitution in the above formula of the coefficients calculated from Ilarrault’s (15) data gives — Trau = .1844, where ran is .2988 36 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE The writer found a correlation coefficient of —.3806 ±.0935 be- tween change in the colony weight and diurnal relative humidity for the spring period. In making a temperature correlation bj7- means of the same formula, he found Tro3 = — .4058. In this particular instance, it appears that dry atmosphere has a beneficial effect upon the change in colony weight. The fall period showed less, or practically no, correlation between changes of colony weight and diurnal average relative humidity, the correlation being — .0960±.1240. It is interesting to note from Tables 2 and 5 that a wide variation of diurnal relative humidity has a beneficial effect upon change of colony weight, and that the coefficients of correlation in this case are higher than those with the diurnal average relative humidity. The condition of the diurnal or nocturnal relative humidity of the day preceding has practically no effect upon change of colony' weight. RELATIVE HUMIDITY AND MIDDAY DECLINE Figures 7, 8, and 10 show that the average minimum relative humidity is reached previous to the average low-point of the midday decline. This fact, together with the existence of a negative corre- lation between net gain and diurnal relative humidity, makes it dif- ficult to assign low relative humidity as the cause for the midday decline. RELATIVE HUMIDITY AND NOCTURNAL LOSS During the spring period the nocturnal relative humidity plays a small part in the nocturnal loss. The coefficients for both the aver- age and the variation in relative humidity are small and have large probable errors. For the fall figures these factors are more important. The coeffi- cient of —.4821 ±-0961 for the average nocturnal relative humidity with nocturnal loss signifies that the drier it is at night the more the*' colony loses in weight. The greater the variation in relative humid- ity during a night, the greater the loss, as is shown by the coefficient of correlation of .5391 ±.0888. Thus it is seen that variation in rel- ative humidity plays a more important part than the average relative humidity upon nocturnal changes in colony weight. SOLAR RADIATION AND SUNSHINE It is a well known fact that both solar radiation and sunshine exert a profound influence upon animal and plant life; it is therefore not surprising to find that these two factors play an important r61e in changes of colony weight. A number of investigators have attempted to learn the effect of light upon nectar secretion, but, as in other experiments upon nectar secretion, various species of plants have been used with the result that authorities differ in their conclusions. Ono (26) placed Ligustrum lucidium, Viburnum japonicum , V. opu- lus, Prunus yedoensis, and P. laurocerasus in a dark moist chamber and found that their young extrafloral nectaries produced no nectar at all, but that their fully developed nectaries secreted in the dark equally as well as in the sunlight. In the case of P. laurocerasus this plant continued to secrete nectar in ample quantities for three weeks while placed in a dark chamber, the nectaries being washed daily. WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 37 He concludes that the influence of light on the secretions of the nec- taries is of an indirect nature, except in the case of Vicia and some species of Euphorbia, these exceptions being noted by Haupt (17). As a matter of fact, Ono concludes that all external factors are of slight value in comparison with the inner conditions of the nectaries. Secretion, he thinks, may occur by mere chance; and the factors which are favorable to the life and growth of the plant at the same time promote the secretion of nectar. Haupt (17) found that for certain plants, e. g., Euphorbia and Vicia, the secretion of nectar is profoundly influenced by light, especially by the red and yellow rays of the sun’s spectrum. The secretion of nectar occurs only in light, and in darkness or in blue light nectaries already containing sugar resorb this substance. Gardiner (12), in commenting on the discoveries of Wilson (33) regarding nectar secretion, says, “I am led to think that this goes far to explain both his own and Darwin’s (6, p. 403) observations, that the exudation of nectar takes place more rapidly in sunlight, for according to my own observations as regards waterglands and the like it does not seem probable that the power of secretion as such is accelerated by light.” However, Wilson (33) does not lay much stress on the effect of light upon secretion, as Gardiner would have one believe, for in Wilson’s conclusions he states that many plants secrete as well in the light as in darkness, while others require either direct sunlight or strong diffused light for secretion. Harrault (14) observed that some plants can yield certain quantities of nectar under weak solar rays but that most plants do not yield nectar even in slight shade. In discussing the influence of external factors on plant transpiration, Livingston, in his revision of Palladin’s work (27, p. 138), says that light is undoubtedly the most important. In the present investigation the total hours of sunshine seem to have a somewhat greater effect upon net gain during the spring period than does solar radiation. The coefficient of correlation of the former with net gain is .6124 ±.0683, while for the latter with net gain it is .5525 ±.0760. In the attainment of the flowering stage of plants, Garner and Allard (13) found that length of day is much more important than the total amount of solar radiation received by the plant. This discovery is interesting in view of the fact that the secretion of nectar accompanies the attainment of the flowering stage for practically all of the important nectar-producing plants. These writers state that — Except under such extreme ranges as would be totally destructive or at least highly injurious to the general well-being of the plant, the result of differences in temperature, water supply, and light intensity, so far as concerns the sexual re- production, appears to be, at most, merely an accelerating or retarding effect, as the case may be, while the seasonal length of day may induce definite expression, initiating the reproductive processes or inhibiting them, depending on whether this length of day happens to be favorable or unfavorable to the particular species. Further investigation along these lines may result in throwing more light on the effect of weather factors on bee behavior, and on plant behavior in so far as the secretion of nectar is concerned. A corre- lation of length of day with nectar secretion of various plants may likewise help to clear up the many differences in plant behavior which beekeepers attribute to differences of locality. 38 BULLETIN 1339, U. S. DEPARTMENT OE AGRICULTURE For the fall period the coefficients of correlation for both solar radi- ation and hours of sunshine with changes of colony weight are prac- tically zero, while negative correlations, having large probable errors, exist between these variables and morning loss. THE EFFECT OF UNKNOWN FACTORS ON CHANGES IN COLONY WEIGHT In this investigation, correlations have been made between changes in colony weight and 10 different factors in the accompanying weather conditions, some of these also being used in correlations with the con- ditions for the day or night previous to the change in weight. There are so many external and internal factors which may influence the physiological processes of nectar-secreting plants and the behavior of bees in gathering and ripening nectar that it is not probable that the factors for which data were obtained are the only ones which may in- The method used in this work is based on Pearson’s theory of mul- tiple correlation. The square of the coefficient of multiple correla- tion between one variable and a number of others measures the extent to which the given variable is determined by the others, while 1 minus this square measures the degree of determination by independ- ent residual factors. In a system of variables X, A, B and 0, in which nothing is assumed of the causal relationships between them, X, which in this case is change in colony weight, is determined by the variables A, B, and C (weather factors) and by all other residual actors. This method has been described in detail by Wright (35). In Figure 11, taken from Wright, the small letters represent the path coefficients, and the capital letters the variables, used in deter- mining the coefficients of correlation. The path coefficients are not equal in value to the coefficients of correlation determined by direct correlation methods, but may be defined as the ratio of the standard fluence the changes in colony weight. For the purpose of ascertaining to what degree the influ- ences have been deter- mined and to what degree unknown factors are involved, an attempt has been made mathe- matically to find the val- ues of all unknown fac- tors. Such a determi- nation throws no light on the actual character of these unknown fac- tors, on the degree of their influence individu- ally, or whether some are positive and some negative, but merely gives the mass value of all factors so far unde- termined. O Fig. 11. — Graphical representation of quantities and their relations involved in the determination of coefficients of correlation WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 39 deviation of X due to each variable to the total deviation of X. 0 represents all residual or unknown factors. From such data we have several simultaneous equations; as many in number as there are variables considered. The following simultaneous equations are obtained: rXA = a +brAB+crAc Txb = CLTab 4* b -f- CTbc rXc = dr AC + brBc + c Fio. 12. — Graphical representation of quantities and their relations involved in the correlation of facto of increase of weight of the colony during the spring and fall honey flows By solving these simultaneous equations the values of the path coefficients a , b, and c are derived. The total extent to which X is determined by the variables A, B, and C is obtained from the follow- ing formula (Wright): 1 - o* = dr ax + brBx + crcx Applying this method to the correlations and other data obtained relating to changes in weight of the colonies during the spring honey flow (Table 2) and those for the fall honey flow (Table 5), ealeu- 40 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE lations have been made by including five of the weather factors. In- Figure 12 the symbols are as follows: G = net gain (changes of colony weight in the fall period), T = average temperature, Tv = temperature variation, H= average relative humidity, IIv = variation of relative- humidity, S = hours of sunshine, L = nocturnal loss, and o = residual, unknown factors. It will be noted that only 5 of the weather factors are here included, although 10 were determined in the original data, but to have included all 10 would have made too unwieldy a calcu- lation. In order to make comparison between the spring and fall periods the same variables were used in both cases. Hours of sun- shine were used instead of solar radiation because this factor has a larger coefficient of correlation with net gain and because there is a high correlation between these two factors of themselves. None of the nocturnal factors were included in this calculation, as being of probably little value in their influence on net gain. For these five weather factors is obtained the diagram shown in Figure 12. From this diagram the following simultaneous equations are- evolved : raT = a + hr TTV -j- CTht + drn vt ~}~ 6T ST Tqtv — CLTttv 4" & 4" CThtv 4~ dVnvTV 4~ 6Vstv Vqii ~CLVth 4~ bVrvH +C + dVnvn + GT sh Vanv = CLVthv -\~^)Vtvhv~\- CVhhv-\- d -\~6Tshv Vgs =” CLTts ~\~1)Ttvs ~\~CThs 4 ~ dTnvs 4“ 6 WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 41 42 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE From values found in Table 2 are derived the following values of the correlations specified, used in calculating the value of the unknown factors influencing change of colony weight during the spring period: Diurnal To • T = .7529 Tt . tv — .5717 TtV . EV — .7855 II 8s .5967 tt-b — — .1551 T TV . s — .7327 To • E — -.3806 Tt • ev — .3899 Te • EV — - .7584 Tg . Ev — .4229 Tt . s — .5798 Tb .8 = -.6420 Tq . s — .6124 T TV • B — — .7113 Tbv • s — .6780 Nocturnal Tl • T — .1754 Tt • tv — — .0739 T TV • B — -.5655 Tl . TV — -.3439 tt-e = — .0043 Ttv • BV — .4493 Tl . b = -.1264 Tt-bv— — .3373 Tb • ev — -.6819 Tl.bv — — .0654 Solving, we find the values of the path coefficients to be — a = .6898, Z> = .0630, c=-.3217, d=~. 2075, e = ’.1004 By knowing the values of the path coefficients, the value of o2 can be determined from the following formula: porT gt 4" PotvT qtv 4~ PgbT gb + PghvT gev 4~ pasT as — 1 — 0 2 and is found to be — o2-. 34705 the degree of determination of unknown factors for the spring period. In determining the value of the unknown factors influencing nocturnal loss during the spring, only four variables were used, namely, nocturnal average temperature, temperature variation, noc- turnal average relative humidity, and variation of relative humidity. The values of the path coefficients were obtained by solving the following simultaneous equations: Tit — CL T T)T'tvt 4 ~ CTbt ~\~cItevt Tltv — CLTttv 4" CThtv T (ITbvtv T lb — CLTtb -\-1)Ttvh 4 ~C 4~ df bvb Tlev — CLT tbv 4- t>T TVHV + ctehv 4- d and are — a= .07329 6= -.58681 e= -.57172 -.16688 The values of the unknown factors are determined by substituting the known values and solving the equation — PltTlt 4- PltvTltv 4~ PlbTlb 4* PlbvTlbv =1 — 0“ whence o2-. 70268 WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 43 It is quite evident in comparing the values of the unknown factors for the diurnal and nocturnal periods that so far as nocturnal loss is concerned the weather factors used have but little influence. The causes influencing the nocturnal loss during the spring period evi- dently lie chiefly within the colony itself. The variables used in calculating the values of the unknown factors during the spring period were also employed in determining the val- ues of the unknown factors for the fall period. The coefficients of correlation, derived from Table 5, and used for this purpose, are — Ditjrnal Tq.t = -.2310 Tt.tv = .0538 Ttv-hv — .6974 To. TV — .5570 Tt.e — .2847 Ttv.s — .5746 Tg.b — -.0960 Tt.bv = — .1184 Tb.bv — -.5995 Tq.rv — .3800 rT.s = .1813 Tb.s — -.6268 To.s = .0595 Ttv.h— — .4095 Tbv.s = .4980 Nocturnal 1! *3 -.3391 Tt.tv— — .1282 Ttv.b — -.4819 Tl.tv — .5136 Tt.b — — .0113 Ttv.ev — .5034 Tl.h — -.4821 Tt.bv— — .1018 Tr.hv — -.9107 Tl.bv — .5391 Substituting the known values and solving the following simulta- neous equations, Tot =CL -\-bvTvr CT et ~^~df,HvT ~h€Tsr Tqtv = (LT ttv ~k & ~k CVjjtv 4~ (Ithvtv 4~ CTstv Tor =&Tth ~k bl'rvH + C ~k dv hvb ~k &T sb T obv — (11'tbv 4~ br tvbv 4~ CT bhv “k d 4~ 6T sbv Tqs = (LTts ~\~1)Ttvs ~\~CVrs -\-dVuvs 4“ 6 we have — a= -.26769, b = . 74349, c = . 16154, d = .04688, e = -.24127 Again, substituting these values for the path coefficients in the following formula, PotTgt 4~ PotvTotv ”k PobT or + paavT obv 4~ pasT as — 1 — 0 2 we find — o2 = . 53609 Although the same diurnal variables were used in calculating the value of the unknown factors for the fall and spring periods, the val- ues of the unknown factors for these periods are considerably differ- ent. There must be other important variables not included in the present data which emphatically influence the changes of colony weight during the fall. The path coefficients for the nocturnal period during the fall were derived from the solution of the following four simultaneous equa- tions: Tit = Cb -\-bTrvT 4 - CTbt 4* dr uvt Tltv — CLTtbv + b "k CTbtv ~k dTuvrv Tin —(LTtu 4* brrvH +c 4- drUvH Tlrv — &Ttuv 4" brrvnv 4" CTi/uv 4" d 44 BULLETIN 1339, U. S. DEPARTMENT OF AGRICULTURE In this case — a= -.27206 c= -.07350 5= ,29408 d = .29642 The solution of the equation — 'PltTlT ’\-rphTvVhTV + 'Plh'TliI + J^LUV r LEV= 1 — °2 therefore gives — o2 = . 56148 It is interesting to note that the value of the unKnown nocturnal factors during the fall is less than it is during the spring. With the same variables in use for calculating the value of o2 during these two Eeriods, the difference must be accounted for by variation in the ehavior of the bees themselves. The writer has calculated the value of o2 from the coefficients of correlation which he determined from data by Harrault (15), given under the discussion of temperatures. In this case only two vari- ables are available, namely, the average daily temperature and the relative humidity. The solution of the two simultaneous equations — Tgt—CL T 1)Tth Tan = CITth + & gives a=.4779 and £>=.1635. By substituting these values in the formula — 'PotTgt + 'PghT on = 1 — O2 we have — - o2= .6998 This figure shows the value of the unknown factors to be greater than that found in this work for the spring period. A similar analysis of Bonnier’s ( 1 ) data gives o2 = .3093. This figure is lower than that given for the spring period. One must take into consideration that only temperature and relative humidity were con- sidered by Bonnier and that activity of the bees does not enter into consideration in his work. THEORETICALLY CHANGING WEATHER FACTORS AND PREDICTING RESULTING GAINS One of the primary objects in presenting the following data is to demonstrate the value of accurately kept records of colony weight, together with weather records. By knowing how changes of colony weight vary with changes in the weather factors over a series of years, it would not be an impossibility to predict whether or not, in the long run, a certain locality would j ustify commercial beekeeping. Such data could also be used to plan migratory beekeeping and to learn perhaps whether a locality not suitable for honey production would be suitable for the production of bees or for queen rearing. In order to give concrete examples in predicting gain under any variation of weather conditions, the necessary data may be taken WEATHER AND CHANGE IN WEIGHT OF BEE COLONY 45 from Table 6 and given, decoded, in Table 7. The following path coefficients are also used in the computations (see p. 42) for the spring period: q/ = 'Pot — .6898 d — jpaHv — — .2075 b = p0Tv = .0630 e = ps = .1004 C — pan — — .3217 The formula for multiple regression, as given below, is used in arriving at the value of the predicted gain G' : 77, T’-T,x Tv' -Tv, H'-H Hv’-Hv,„ S'-S G —6 r + fl(T(j —-\-b(Ta -j-Ccra h da a r^c To- ff T . 611-012. 44349°— 25 1 1 2 BULLETIN 1346, U. S, DEPARTMENT OF AGRICULTURE State of Mexico. The plain on which this hunt occurred has been known from that day to the present as the Llano del Cazadero, in commemoration of this event. The station of Cazadero on the main, line of the Mexican Central Railroad marks this vicinity. Fig. 1. — Original and present distribution of the pronghorned antelope. The black line marks the limits of the distribution before European, settlements in America. The shaded portion indicates the area within which antelope are now found (192,2—1924) in, scattering bands. Details of present distribution within this area are shown on separate maps of States and Provinces (figs 3 to 21), and in Table 1, page 3 The hunt took the form of a great drive of game by the Indians, during which, the author states, 600 deer were killed, among which were large stags “ and those which they call verrendos.” He states that the verrendos did not occur in Spain, and that “ they not only ran but flew,” thus indicating that the remarkable speed of these STATUS OF THE PRONGHORNED' ANTELOPE, 1922-1924 3 animals attracted the attention of the first European observers. Throughout the antelope country in Mexico and the southwestern United States the Mexicans still term these animals “ berrenclos,” the “ v ” of the old Spanish having been replaced by the modern “ b.” As a matter of course the pronghorn must have been a familiar animal to the hardy Spaniards, who overran all parts of Mexico and much of the southwestern and western United States in their search for gold, but their records of the animal life seen are exceedingly scanty. Subsequent occupation of the continent has shown that the prong- horn ranged over an enormous area. (See map, fig. 1.) It occurred over parts of the present Provinces of Manitoba, Saskatchewan, and Alberta, in Canada. In the United States it occupied the country from western Minnesota, Iowa, Kansas, Oklahoma, and Texas, reach- ing the Gulf coast near the mouth of the Rio Grande, and Avest to eastern Washington, Oregon, and the Pacific coast in California. In Mexico it occupied the open plains country of1, the tableland south almost to 20° of latitude, nearly to the Valley of Mexico; also the western part of Sonora and most of Lower California. Table 1. — Distribution of antelope in North America, 1922-1924 Region Areas Number of ante- lope Region Areas Number of ante- lope Arizona 18 651 Canada: California 6 1, 057 5 1, 030 Colorado ... 28 1, 233 9 297 14 1 486 Kansas i 8 Total, Canada 14 1, 327 Montana 44 3,027 Nebraska 10 187 Nevada 11 4, 253 1 31 1,682 Durango (?) Sonora 4 595 2 23 Lower California 1 2 500 Oregon. 4 2,039 Total, Mexico 8 i 2, 395 11 680 Texas... 42 2, 407 Summary: i tab 10 670 264 26, 604 Wyoming 27 6,977 14 1, 327 8 1 2, 395 264 Grand total _ _. 286 30, 326 1 Estimated. Through the occupation of its territory by man the pronghorn has become extinct in many of its former haunts, but it has survived in limited numbers over an amazing proportion of its original range in Canada, Mexico, and in 16 of the western States of this country. Originally over most of the enormous territory occupied the prong- horn was very abundant. Its range covered not only practically all of the buffalo country west of the Mississippi River but a vastly greater area. Where the pronghorn occurred with the buffalo people best qualified to judge consider that it exceeded that animal in num- bers. It has been estimated that the buffalo herds at one time num- bered from thirty to sixty million animals. In view of the greater territory occupied by the pronghorn and its known abundance, it may be considered a conservative estimate to place its probable origi- 4 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE nal numbers at not less than thirty to forty millions, and possibly more. George Bird Grinnell informed the writer that he has often talked about the abundance of antelope with men familiar with the western plains 50 years or more ago and has never met a man of experience who did not agree with him that during the middle of the last cen- tury antelope were far more abundant than buffalo. During the summer of 1879 Doctor Grinnell found them extremely abundant in North Park, Colo., where he saw trails made by them in travel from one locality to another worn in the hard soil to a depth of from 8 to 10 inches, like the trails made by buffalo herds going to and from water or during their movements from one district to another. As against the many millions of pronghorns once inhabiting this continent a recent census, taken through the Biological Survey and detailed elsewhere, shows approximately 30,000 survivors. (See Table 1, p. 3.) CHARACTERISTICS OF THE AMERICAN ANTELOPE Horns. — The pronghorn is the only antelope in the world with branched or pronged horns, and has the unique characteristic among all hollow-horned ruminants of shedding the outer covering of the horns annually. This takes place soon after the rut in November and December in the Yellowstone National Park in northern Wyo- ming, and elsewhere in the range of the species this time probably varies somewhat with latitude. When the time for shedding arrives the horny sheath gradually loosens and becomes detached from the skin around the base and, following this, from the bony core within. Later the horn falls off, leaving the bony core covered with a blackish skin more or less overgrown with long, coarse hairs, which afterward are gradually lost. A new horny nucleus develops on the tip of the bony core, the horny growth then extending slowly downward until it reaches the base. Gradually thickening and hardening, the horny material grows at the tip until the new horn attains its full development. The horns continue to grow as the animal increases in age until the full size is reached. Both sexes have horns, those on the does being smaller and slen- derer than on the bucks. Rump patch . — Another characteristic of these interesting animals is a conspicuous rump patch composed of white hairs which are longer than those elsewhere on the animal’s back. Through develop- ments in the skin muscles the pronghorn at times of excitement has the power to erect these white hairs until they stand out stiffly over the rump, forming a great dazzlingly white rosette, like a giant chrysanthemum, which, when the animal is dashing away across the plains in the bright sunlight, is extraordinarily conspicuous. The writer has many times discovered bands of antelope running on the open plains where but for these heliographic patches they would have been beyond ordinary eyesight. These long rump hairs lie like other hairs on the skin and give little indication of their strikingly conspicuous appearance until the animal suddenly throws them up into action. The antelope fawns at a very early age STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 5 begin “ flashing ” their white rump patches on being startled or excited. Curiosity. — In addition to its physical peculiarities the pronghorn is very different psychologically from any other of our large-game animals. Early in their acquaintance with these animals hunters became familiar with their intense curiosity, and have employed various methods to toll them within gunshot. One of these was to lie on the ground and wave a red flag slowly back and forth on a ramrod. Another strange performance often said to have the same effect was for a hunter to lie on his back and kick his heels in the air. While in Mexico some years ago, during the Biological Survey zoological explorations, the writer located a considerable number of antelope on the grassy plains of northwestern Chihauhua, but found them so shy from being hunted in these open spaces that they were almost impossible of approach within gunshot. Specimens were needed for the bureau’s scientific study series and every effort was made to secure them — at first, owing to the shyness of these animals, almost without success. Finally, recalling old stories of the curiosity of the antelope, the writer tried the experiment of taking a white bed sheet and, placing one edge over his head, fastened it under his chin. This formed a kind of hood, and when the two upper corners were passed under his arms and attached at the middle in the back, and the hanging edges fastened in front of his body, the whole formed a kind of hooded cloak completely covering him from head to foot. A lot of long grass stems were then gathered and stuck through his hatband so to form a tall, grassy crown. Covered with this white cloak the writer rode out on the plains until he located a band of antelope, and when at a distance of nearly half a mile dismounted, hobbled his horse, and proceeded toward them in a stooping posture. Meanwhile they were standing look- ing fixedly at him. When he came within 500 yards he vrent on his hands and knees, the sheet covering him to the ground, and began moving slowly toward them. The antelope had lined up, with a large buck standing in front. They turned several times and nervously ran a short distance and then turned and raced back to their first position, where they lined up to look at the strange object. The old buck of the band, which from the beginning had stood out by itself in front, began slowing walking toward him. The writer then stopped and sat with crossed legs, the cloak still hiding his person, and waited, rifle in hand, until the buck had come within 100 yards, when it became a prize for the bureau’s scientific collection. On another occasion, while clad in the same disguise, the writer saw a solitary old buck antelope standing about half a mile away on the far side of a bare, dry, alkali mud flat. He again dismounted and made a similar approach, the buck meanwhile standing and Avatching him steadily. The buck remained motionless and per- mitted the writer to continue to approach until within about 100 yards without showing any sign of alarm. Another solitary buck grazing on an open grassy plain was ap- proached in the same manner. Whenever it' stopped grazing and looked at him intently, the writer moved his head up and down and 6 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE sidewise as though feeding on grass and looking about and then continued to advance on hands and knees. Finally the animal stopped grazing, and when the writer was well within 100 yards it actually closed its eyes and appeared to be dozing, as its head nodded slightly up and down, apparently in complete indifference. Through the use of this sheet the writer had no trouble in ap- proaching antelope anywhere on the plains, and he was inclined to think that they took him for some harmless white animal. There were many half-wild cattle grazing on these plains at that time which were ordinarily shy and would run away when a man ap- peared on horseback. After the writer began wearing this white sheet, which not only covered himself but when on horseback would spread over the rump of the horse, the cattle ceased to show any alarm as he appeared and would permit him to ride through herds of them, merely lifting their heads and gazing at him for a short time and then resuming their feeding, the effect of the disguise ap- parently being the same with them as with the antelope. It may be of interest to know that the specimens of antelope secured by the use of this grotesque disguise formed the basis of Doctor Merriam’s description of a new geographic race of the pronghorn, which he named Antiloca-pra americana mexicana.2 Racing. — One of the most extraordinary peculiarities in the psychology of the pronghorn is its desire to pass in front of a mounted man or a team moving by at no great distance from a band. From 1883 to 1888 the writer lived in a section of Arizona where antelope were plentiful, and frequently hunted them and often saw them when riding in a wagon or on horseback along roads or trails crossing their haunts. This area was mainly covered by a great scattered forest of pinyons, cedars, and junipers, interspersed with many small grassy parks of varying size. During the summer antelope were distributed in small bands in these parks, sometimes 2 or 3 individuals together and at other times from 15 to several times that number. It was a common occurrence when a traveler passed along these roads for a band to stand from 75 to 200 yards away watching him. Then they would suddenly start and run one after the other parallel to the course taken by the traveler and dash across the road immedi- ately in front of him, often within a short distance, after which they would stream away and disappear among the scattered .tree growth. When traveling on horseback and happening upon antelope in such places the writer often amused himself by spurring his horse to a gallop and continuing his course in a direction which would take him by and away from the animals. At such times he tried to appear un- conscious of their presence, and this procedure almost invariably brought the expected response, and the animals began racing him until they had gained a slight leadership, when they would dash by in front across the road or trail, one after the other, frequently the last of the lot being within 20 yards. Once the writer tried the experiment when he saw a solitary buck antelope stand about 100 yards to one side of a wagon road. Ap- pearing not to notice it, he spurred his horse at full speed across 2 Proc. Biol. Soc. Washington, vol, 14, p. 31, 1901 ; type from Sierra en Media, Chihua- hua, Mexico. STATUS OF THE PEOHGHOBNED ANTELOPE, 1922-1924 7 the level plain. The buck immediately whirled and began racing him over the grassy park, gradually drawing in until it finally crossed the trail almost under the horse’s nose and certainly not more than 10 feet away, after which it dashed off and disappeared in the neighboring scattered growth of cedars. On another occasion, after a long hunt, the writer was returning to camp just as it was becoming too dark to distinguish objects at a distance. Camp was some miles away, and in order to get there .quickly he was galloping his horse down the middle of a long, narrow park in the scattered pinyon and cedar forest. He was paying no attention to anything except what lay immediately in front until a curious sound at his right caused him to look, and there he made out the dim forms of a band of about 20 antelope which, in a long line about 30 yards away, were racing him down the park. Eventually they gained sufficient headway to cross his course a short distance in front, when they disappeared. It was so dark at the time that their forms could be only dimly seen. In discussing the pronghorn with many hunters who have been familiar with it in early days the writer has noted that without ex- ception they have accounts illustrating the extraordinary and appar- ently overwhelming curiosity of these animals. This very fre- quently has led the animals to expose themselves to the most immi- nent danger. They sometimes would come almost into the midst of a camp to satisfy themselves as to the strange beings who had suddenly appeared in their territory, and many fell victims to this habit. CHOSEN HABITAT ■ The natural home of the pronghorn was on the treeless, grassy, and often desert plains of the continent. The animals would scatter singly or in small bands in spring and summer, especially during the period when the does were caring for their young fawns. As winter approached they began to gather in bands, sometimes con- taining thousands of individuals, and to seek favorable feeding grounds for the winter. A band of more than 500 frequented a broken and open pinyon and cedar forest in the part of eastern Ari- zona where the writer lived in the early eighties. In summer they broke up and scattered over the more open plains in the adjacent parts of New Mexico and northern Arizona. Numbers of them con- tinued to reside through the year among the pinyon and cedar forests, but the bulk of the band went out on the grassy plains. In winter they were very fond of gathering in the pinyon and cedar forests, where they were sheltered from the cold storms which made the open plains places of discomfort. When within these scrubby sheltering forests they were especially liable to become victims of predatory animals and hunters. Near the base of the Sandia Mountains, in New Mexico, the writer knew of hunters trailing bands of antelope among the pinyons during long-continued snow storms and killing many of them one after the other. The animals thus falling victims to the hunter would be roughly dressed and hung up in a pinyon tree, and then the hunter would resume the trail of the survivors and in a comparatively short distance again overtake them and obtain another victim. In this way as many as 10 or 12 could bo killed at times during a single morning. 8 BULLETIN 1346, U S. DEPARTMENT OF AGBICULTURE During the eighties the increase in the cattle business was so great in northern Arizona that the antelope learned many new habits. Among others was that of following range cattle through a belt of heavy pine forest up to an elevated grassy plateau of about 8,000 feet altitude, lying on the east front of the White Mountains, about the headwaters of the Black, Blue, and Colorado Rivers. There, on a wide rolling open plain, they passed the summer, coming out, on the approach of winter, in company with the cattle. This change was comparable to that which caused the elk, once a habitant of the foothills and adjacent plains, to become an animal of the higher elevations. During this period antelope became frequenters of the open, grass-grown, yellow-pine forests of the mountain areas not only in various parts of the United States but also in the Sierra Madre of Chihuahua, Mexico. CONSERVATION AND CONTROL The hunting of antelope is now forbidden by law almost through- out its range. In the United States, of the 16 States in which these animals still occur, Wyoming is the only one in which their hunting might be legalized. The Wyoming law authorizes the State game and fish commission to permit the killing of not to exceed 100 bucks in designated parts of the State from September 15 to October 31 in any year. In 1922 the Wyoming commission had in mind to permit the killing of 100 buck antelope under the terms of this law, but the opposition expressed by individuals and in the press, not only in Wyoming but in other parts of the country, caused the plan to be abandoned. In 1925, however, the legislature authorized the issuance of 300 such permits during October, in certain counties in the eastern and southern parts of the State. In Nevada the close season ends in 1930, and in Kansas, by action of the 1925 legislature, the close season was extended indefinitely. There is little likelihood that the season will be opened in any other States in the near future, although under good protection the increase of antelope in favorable areas may in a few years render it urgently necessary to reduce their numbers. With the increasing occupation of the western United States, the presence of antelope in such numbers as might occur under complete protection might create a situation that would be intolerable to some of the residents whose livelihood depends upon farming and grazing. Antelope, as in the case of other large-game animals, when under practically complete protection, lose their fear of man to a surpris- ing extent and become bold in raiding fields and in destroying crops. The possibility of the development of such conditions should be seriously considered by conservationists in building up herds of antelope. Efforts should be made to seek, for the establishment of antelope refuges, remote and thinly settled areas unless the animals are to be reared within fenced inclosures. Even in the latter case the increase of the animals will eventually require some control of the numbers by eliminating the surplus. This is a matter of prac- tical game administration which should be understood and accepted by the public with the same matter-of-course attitude that is shown toward the control of the surplus livestock on a farm. The limited hunting-license system provides a practical method of handling sur- Bui. 1346, U. S. Dept, of Agriculture Plate 1 1 B2499M Fig. I -Two Newly Born Antelope, Diessner Ranch, Nev. The doe usually hides the fawns one in a place and, feeding unconcernedly, approaches to nurse them. By watching the mothers the young animals are located and captured B2G02M Fig. 2. -Fawns Just Captured, Washoe Rufuge, Nev. As soon as captured, the fawns are placed in grain sacks, each with a hole in it just largo enough for the head to be put i h rough. One sack is then hung on each side of a saddlo horse and the fawns arc brought to camp Bui. 1346, U. S. Dept, of Agriculture Plate III B2505M Fig. I.— Young Antelope in Temporary Inclosure Captured on the Washoe Game Refuge and raised on the Wood Ranch, near Diessner, north- western Nevada, by E. R. Sans, of the Biological Survey. Photograph taken July IS, 1924 _ B2503M Fig. 2. ^Bottle-Raised Fawns Photograph taken in August, 1924, on the Wood Ranch, near Diessner, Nev. STATUS OF THE PEONG-HORUED ANTELOPE, 1922-1924 9 plus game and at the same time perpetuating the species in reason- able numbers. Conservationists should appreciate that there must be a sympa- thetic attitude on the part of the general public which is in direct contact with the game. This means that the surplus of big game, either of antelope or of any other kind, must be disposed of in a practical way, leaving a reasonable breeding stock to perpetuate the species. Those having administrative charge* of the game in any area should determine the number of each species of large game that can properly be maintained there and provide for eliminating the surplus, if any, each year. In this manner overstocking the range and other embarrassments may be avoided. The number of game animals of each species to be maintained in each area should be determined by a careful study of the conditions in the area by trained experts having practical knowledge of the requirements of the different species. It should be clearly under- stood that at the present time and for some years to come antelope need strict protection practically throughout their range, the one probable exception being in a limited area in Wyoming. CONSERVATION ORGANIZATIONS AND THE ANTELOPE A number of leading organizations of the country have interested themselves definitely in the conservation of antelope during a period of years. This is an appropriate place to put on record the more notable achievements of these organizations up to this time. Boone and Crockett Club. — The last of December, 1910, and the first of January, 1911, 3 buck and 6 doe antelope were obtained from the Yellowstone National Park herd by the Boone and Crockett Club and shipped to the Wichita National Came Preserve in Okla- homa, in cooperation with the United States Forest Service. (See PI. I.) During this period 4 bucks and 8 does from the same source were shipped by this club to the National Bison Itange in Montana, in cooperation with the Biological Survey. Of the antelope sent to Oklahoma, some reached their destination dead and others badly in- jured, and eventually all died. Those sent to the Bison llange were the basis of a herd of 64 animals which was subsequently built up and then completely destroyed by the inroads of predatory animals. During 1914 the Boone and Crockett Club purchased 13 antelope in Alberta, which were sent to the Wind Cave National Park and Game Preserve in South Dakota. These animals did well for a time, but later their numbers had become so reduced that in 1916 the club again purchased 9 antelope in Alberta, which were placed in the same game preserve. Here they did well for a time and in- creased to 34 animals. Most of these were afterwards destroyed by predatory animals, but enough still remain to serve as a nucleus for building up a new herd. Special efforts are being made to protect them from further attacks by predatory animals. (PI. VI.) The Boone and Crockett Club appears to have been the pioneer in this line of conservation. American Bison Society . — In January, 1912, the American Bison Society first became interested in the future of the antelope and assisted in an attempt to pass a bill through Congress to create the 44.349°— 25 2 10 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE Snow Creek Antelope Preserve in Montana. In 1919 M. S. Gar- retson, secretary of this society, accompanied a representative of the Biological Survey to investigate conditions in southeastern Oregon for the purpose of determining the suitability of the region for the establishment of a Federal antelope refuge. In October, 1921, the society purchased in Alberta and delivered at the Wichita National Game Preserve, Okla., 10 young antelope for the purpose of again trying to establish a herd in this locality. Most of these perished, and in 1922 the society purchased 6 addi- tional antelope in Alberta, which were delivered at that game pre- serve. Most of these died ; but, as is set forth in detail in the account of antelope in Oklahoma, this herd is at last increasing, with a good prospect that it may become permanently established. During the spring of 1921 the secretary of the society accompanied a representative of the Biological Survey on an expedition to south- western Idaho to examine that district to make recommendations concerning the establishment there of the Owyhee Antelope Befuge. The same year funds were contributed by the society to assist in protecting the Mount Dome antelope herd in California. The annual report of the American Bison Society for 1922-23, pages 49 to 51, contains the first published census of the antelope of North America, which is dated January 1, 1922. This census was compiled by M. S. Garretson, secretary of the society, and gave a total of 11,749 antelope for the United States and Canada. The census of antelope compiled by the Biological Survey and given in detail later in this publication indicates a much greater number of antelope surviving than given in the first census mentioned above. Mr. Garretson’s very creditable work was handicapped by many difficulties. The greater completeness of the census of the Biological Survey is clue to the fortunate fact that this bureau has a number of employees permanently located in each of the several States where antelope occur. The Bison Society has taken the stand that, having assured the perpetuation of the bison, it is free to help save the antelope now approaching a condition which will require active work to prevent their extermination. Permanent Wild-Life Protection Fund. — The Permanent WiicI- Life Protection Fund throughout its existence has taken an active in- terest in the conservation of antelope and has contributed substantial sums to carry out this purpose in various parts of the country. It contributed to the Bison Society funds in connection with investiga- tions to create a Federal antelope refuge in southeastern Oregon, also for the establishment of a herd of antelope in the Wichita Na- tional Game Preserve, and for the protection of the Mount Dome herd in California. In cooperation with E. E. Brownell and the Biological Survey it assisted in the capture of 40 young antelope in northwestern Nevada during the spring of 1924 and in their distribu- tion to the Grand Canyon National Park, Ariz. ; to the Niobrara Federal Game Befuge, near Valentine, Nebr. ; and to the National Bison Bange, in western Montana. It has also contributed funds to cooperate with the Biological Survey in marking the boundaries of the Washoe and Humboldt Antelope Befuges in northern Nevada. It has contributed special rewards for convictions for illegal killing of STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 11 antelope in Oregon, and is cooperating with the Mexican Govern- ment in maintaining a special warden service to protect antelope and mountain sheep in northern Sonora. There the Mexican Government has appointed Ben H. Tinker, of Arizona, an honorary game guardian for northern Sonora. He entered on this duty on October 1, 1923, and patrols the Arizona-Sonora border during the active hunting season, October 1 to April 1, in order to enforce an executive decree protecting antelope and mountain sheep for a period of 10 years. California Associated Societies for the Conservation of Wild Life. — This organization has interested itself in the perpetuation of the Mount Dome antelope herd. It is working with the State board of fish and game commissioners and has provided funds for feeding the animals during severe winters. It began its operations in 1914,' and among its other activities materially helped in gathering infor- mation for this report concerning the distribution of antelope throughout California. WASHINGTON CONFERENCE ON THE CONSERVATION OF THE PRONGHORN For some years suggestions had been made that a conference be held to consider the conservation of antelope, the earlier proponents being E. Lester Jones, Director of the Coast and Geodetic Survey, Department of Commerce; Edmund Seymour, president of the Amer- ican Bison Society; T. Gilbert Pearson, president of the National Association of Audubon Societies ; and others. In view, however, of the fact that a census of these animals was being taken by the Biological Survey, it was deemed best to delay such a conference until this investigation had been concluded, in order that the information obtained might be available for consid- eration. This census Avas practically completed in the fall of 1923, and a call for the antelope conference was issued by the chief of the Biological Suiwey to meet in the auditorium of the NeAv National Museum, in Washington, D. C., December 14, 1923. The meeting was attended by representatives of the principal wilcl-life-conservation organizations of the country, State game wardens from a number of States, representatives of Government bureaus interested in Avild- life conservation, and numerous private individuals. The conserva- tion organizations and their representatives Avere: Boone and Crockett Club, by Charles Sheldon; American Bison Society, by Edmund Seymour and W. T. Hornaday; Permanent Wild Life Protection Fund, by W. T. Plornaday; National Asso- ciation of Audubon Societies, by T. Gilbert Pearson, W. P. Whar- ton, and William Finley; American Game Protective Association, by John B. Burnham; Izaak Walton League, by Will PI. Dilg; Game Conservation Committee of the Camp-Fire Club of America, by W. B. Greeley and Marshall McLean ; Associated Societies for the Protection of Wild Life in California, by Alden Sampson; and the National Parks Association, by Robert Sterling Yard. Registration was made of the following State game commissions, represented by their chiefs: Arizona, G. M. Willard; Arkansas, Lee Miles; California, F. M. Newbert; Kansas, J. B. Doze; Massachusetts, W. C- Adams; Minne- 12 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE sota, J. F. Gould; Pennsylvania, the executive secretary, Seth E. Gordon. The interest of the Federal Government in the conservation of the pronghorn was shown by the presence of Senator Peter Nor- beck, of South Dakota, and by representation from the following Federal bureaus: National Park Service, the director, Stephen T, Mather; Forest Service, W. C. Barnes and E. 1ST. Kavanaugh; Coast and Geodetic Survey, the director, E. Lester Jones; and the Biological Survey, by the chief and several members of the staff. The Canadian National Parks Service, at the request of J. B. Harking commis- sioner, was also represented, O. S. Finnie, director, Northwest Terri- tories Branch, Department of the Interior, being present. The results of the census which had been practically completed by the Biological Survey were presented. These are set forth with some additions elsewhere in this bulletin. From time to time during the past years the desirability has been suggested of organizing a national antelope society to foster the conservation of the pronghorn. In view of the fact that a number of the more important conservation organizations had already in- terested themselves and expended considerable money on projects for this purpose, it appeared to the conference that the organiza- tion of an additional conservation societj^ would probably result in complications not beneficial to the cause and might really act as a deterrent to the development of much-needed activity to save these beautiful animals from extermination. It was agreed that the existing conservation organizations should continue to interest themselves in the conservation of the antelope as opportunity offered, and that whenever one organization should take up a specific project the others would actively cooperate in carrying it to a successful conclusion. It was further agreed that the Biological Survey should serve as a clearing house for information concerning the pronghorn, and that its cooperation should be utilized as fully as possible in this work. The location of the Biological Survey field men engaged in predatory-animal and rodent-control work in all of the States in which antelope still exist places this bureau in a specially favorable position to procure up-to-date information on the subject. To put this decision into definite form Marshall McLean proposed a reso- lution for the purpose of establishing continuity of interest and activity, which was unanimously adopted by the conference, as follows : That individuals and organizations represented here or others desiring to take part constitute themselves a conference for the preservation of antelope with the object of cooperating with the Biological Survey to that end. After a discussion of many details and phases of the antelope situation and of matters connected with their conservation, T. Gilbert Pearson presented L. D. Frakes, owner of a cattle ranch near Warner Lake, and J. L. Lyon, owner of a sheep ranch near Lake- view, in southeastern Oregon, who came to the conference for the purpose of advocating the establishment of an antelope and sage-hen refuge covering a large area in southeastern Oregon. They an- nounced that their ranches lie within the proposed refuge and that STATUS OF THE PRONG HOUXF.D ANTELOPE, 1922-1924 13 up to within a year they had been strongly opposed to its establish- ment; but that, after studying the matter carefully and learning the facts as to the policies that would be followed by the Govern- ment, they had become convinced that such a refuge would be to the advantage of the residents of that section of Oregon, as well as to the antelope and other wild life there. In the discussion with these stockmen it developed that in the period when they were opposing the establishment of the refuge they believed that it would involve the elimination of livestock within the area and the destruction of their interests; but when it was understood that if such a refuge should be established there would be no elimination of the livestock of resident stock growers or other interference with their freedom beyond that of stopping the shooting of game within the area and the limitation of the num- ber of livestock to the capacity of the forage production of the range, they approved the plan. After discussion of the suggested antelope refuge in southeastern Oregon the conference adopted a resolution authorizing a committee made up of representatives of the conservation organizations present to meet in the offices of the Biological Survey during the afternoon of December 14 to confer with those interested and draft a bill for the creation of an antelope and sage-hen refuge in southeastern Oregon. Since then a bill has been introduced in the Oregon Leg- islature for the establishment of a State antelope refuge covering the area recommended. The exclusion of hunters from the area and the prevention of de- structive overstocking cover the only restrictions contemplated in the proposed Federal refuge. It was planned that the control of the grazing should be under the supervision of the Forest Service, in order to provide for the best utilization of the forage practicable. It will be of interest to know that the first suggestion for estab- lishing an antelope refuge in southeastern Oregon appears to have been made by L. Alva Lewis, an agent of the Biological Survey, in a letter dated January 22, 1913. In October of the same year Harry Tilford, inspector of State game refuges for the State Game Com- mission of Oregon, made a similar recommendation. In 1916 E. Lester Jones, Director of the LTnited States Coast and Geodetic Survey, made a trip into eastern Oregon, where he ob- served the antelope in the vicinity of Desert Lake, Jacks Lake, and Guano Lake, the principal herd being in the vicinity of Jacks Lake and containing about 800 animals. In all, he saw more than 1,000 antelope, including a number which had been apparently wantonly killed and then left to lie undisturbed where they fell. On his return from this trip Colonel Jones advocated the establishment of a Federal antelope refuge in Lake County, Oreg., in order to try to prevent the destruction of these herds, and at the National Parks Conference held in Washington, January 4, 1917, he delivered an ad- dress on “ The future of the antelope,” which was printed with a map showing the proposed refuge and distributed as a circular by the National Park Service. His recommendation included the country from Hart Mountain east to the Lake County line and south to in- clude Guano Lake. Colonel Jones warmly advocated the holding of 14 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE the antelope conference, and still maintains a keen interest in an- telope conservation. In 1917 and 1918 representatives of the Biological Survey investi- gated and reported on the antelope situation in that region with a view to the possible establishment of a Federal refuge. In 1919 the secretary of the American Bison Society visited the area in company with a representative of the Biological Survey, and later recom- mended to his organization that it interest itself in the establishment of the proposed refuge. Ever since the refuge was first proposed the National Association of Audubon Societies, through its Oregon representative, William L. Finley, has taken an active interest in it; and in 1923, through the efforts of the Biological Survey and the National Association of Audubon Societies, a meeting of stockmen of Lakeview was held, and favored the creation of a Federal wild-life refuge, with the understanding that local stockmen within the area should continue to retain their grazing and other rights. It was as a result of this meeting that resident stock growers attended the antelope conference in Washington, as set forth above. ESTABLISHMENT OF ANTELOPE REFUGES IN NEVADA In 1923 Gov. J. G. Scrugham, of Nevada, who had been empowered by the State legislature to create 25 State game refuges, requested the assistance of the Biological Survey in determining suitable loca- tions, particularly those to be made for the protection of antelope. E. R. Sans, supervisor of predatory-animal and rodent-control work of the Biological Survey in the State, consulted with the governor, and as a result two antelope refuges were established — the Washoe State Recreation Ground and Game Refuge, lying mainly in Washoe County, in extreme northwestern Nevada, adjoining Oregon (PI. V, fig. 1) ; and the Humboldt State Recreation Ground and Game Refuge, on the northern border of the State, adjacent to Owyhee County, Idaho. Following their establishment, on recommenda- tion of Mr. Sans, who was familiar with the region, enlargements of these refuges were made by the governor to include adjacent districts specially frequented by antelope herds. At the time of its creation the enlarged Washoe refuge was be- lieved to contain about 2,000 antelope, and the Humboldt refuge about 1,000. The Washoe refuge lies adjacent to that part of south- eastern Oregon which for some years has been under consideration as the possible site of a Federal refuge for antelojoe and sage hens. The Humboldt refuge lies immediately to the south of that part of southwestern Idaho which also has been considered as a possible Federal antelope and sage-hen refuge. The establishment of refuges in adjacent parts of Oregon and Idaho would thus afford protection to the antelope herds passing back and forth across the border in this great tableland region, which is obviously so favorable to the perpetuation of these animals. Predatory-animal hunters under Mr. Sans’s direction had been working for a long time destroying cojmtes and other stock- and game- killing animals in the region covered by the Nevada State antelope refuges. Work for the destruction of predatory animals in these STATUS OF THE PRO 1ST Gif 0 RNE D ANTELOPE, 1922-1924 15 areas will continue and will be a great factor in lessening the losses of young antelope and in building up the herds. In addition, the Biological Survey has made an exception to its general rule and has permitted its hunters to be made deputy State game wardens, so that in carrying out their predatory-animal-control work they will be in position still further to assist in the protection of the antelope herds. Local stockmen have shown a most friendly spirit toward the establishment of the Washoe and Humboldt refuges and have ex- pressed a desire to assist in the protection of the antelope within these areas. The Washoe refuge contains about 3,888 square miles, and its boundaries are about 312 miles in extent. The Humboldt refuge covers an area of 1,836 square miles, with a distance of 168 miles about its borders. At the request of the governor, the Biological Survey is taking charge of marking the boundaries of both refuges. This is rendered possible through a generous contribution of funds from the Permanent Wild Life Protection Fund, through W. T. Hornaday. Metal signs bearing the following legend are being placed on posts at suitable intervals around the borders of the Washoe refuge, and similar signs about the Humboldt refuge : NEVADA GAME REFUGE NO. 9 Foe the Preseevation of Antelope and Other Game HUNTING GAME ANIMALS OR BIRDS ON THIS REFUGE IS PRO- HIBITED UNDER PENALTIES PROVIDED BY LAW All persons are asked to assist in the protection of antelope, to prevent the extermination of this beautiful animal, found only in North America ; also to help protect other game, that the surplus may spread to the surrounding country. Maintained in cooperation between the State of Nevada ; the Bureau of Biological Survey, United States Department of Agriculture; and the Permanent Wild Life Protection Fund of New York. For further information address Nevada State Game Commission, Carson City. J. G. Scrugham, Governor. PROPOSED OWYHEE ANTELOPE AND SAGE-HEN REFUGE, IDAHO Apparently the first suggestion that a refuge should be made for antelope and sage hens in southwestern Idaho was in a letter dated December 10, 1920, from George Tonkin, United States game warden in that region. In 1921 further information was received from Mr. Tonkin and other representatives of the Biological Survey in that area, and in the same year the American Bison Society became interested in the project, and its secretary, M. S. Garretson, visited the Owyliee region with a representative of the Biological Survey. As a result of his report and recommendations the Bison Society became active in trying to bring about the establishment of this refuge. The stockmen resident within the limits of the proposed refuge in Owyhee County were practically a unit in its favor on the 16 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE understanding that those operating within its limits should retain their existing rights. This, as in the case of the proposed antelope refuge in southeastern Oregon, accorded perfectly with the policy of the Biological Survey and the desires of its cooperators inter- ested in the project. Many antelope now exist within the limits of the proposed Owyhee refuge, as well as mule deer and other interesting mammals, and many sage hens. It is an ideal arid-region game refuge, offering sage plains varied with groups of low mountains, where consider- able numbers of game animals can be maintained without in any way interfering with the grazing interests. The sympathetic interest shown by the stockmen in this area indicates that if this refuge can be established it will give a fine demonstration of the practicability of maintaining reasonable num- bers of game along with the continued use of such an area for stock- growing purposes. It can not be too often emphasized that it is not the policy of the Biological Survey completely to exclude grazing from game refuges except under very exceptional circumstances. This bureau is convinced that wherever a large area is involved game can be maintained there with stock without interfering with the legitimate utilization of such area for economic purposes. It is hoped that with a better understanding of the purposes of these refuges, which are mainly to put an end to hunting game within their limits and to bring about a greater protection of game in order that it may be maintained and increase for the benefit of the surrounding region, the neighboring stockmen may come to approve their establishment. It is to be appreciated that refuges of this character should be formed in complete cooperation with the stockmen if they are to be effective. RESTOCKING EXPERIMENTS, 1924 After a visit to the Grand Canyon National Park in northern Ari- zona, E. E. Brownell, of San Francisco, suggested in 1922 that the plateau midway down the slope on the south side of the canyon might well be utilized to establish a band of antelope. Following this suggestion, an expert of the Biological Survey examined the ground and found the project to be practicable. W. T. Hornaday, of the Permanent Wild Life Protection Fund, also visited the Grand. Canyon and approved the establishment of an antelope herd there. The outcome was that Doctor Brownell and Doctor Hornaday each contributed a very considerable sum to a fund to cooperate with the Biological Survey to carry out the project. At first it was planned to purchase the young antelope for re- stocking purposes from Alberta, but later the Governor of Nevada, in recognition of the cooperation of the Biological Survey in the establishment of the State antelope refuges in the northern part of the State, very generously granted a permit for the bureau to cap- ture 40 antelope fawns on the Washoe State Game Refuge. For- tunately O. C. Wood, one of the predatory-animal hunters of the Biological Survey, owned a ranch in the midst of the area occupied by the great antelope herds of that region, which was an ideal place Bui. 1 346, U. S. Dept, of Agriculture Plate IV _ _ B25 1 7M Fig. I.— Antelope Fawns in Inclosure Photograph taken September 4, 1924, in a wire-fenced inclosure built for captured fawns, adjoining a house in Reno, Nev., where the young antelope were held for one month wTith complete success 13251 6M Fig. 2. Young Buck and Doe Two members of the band of fawns shown above. The doc, about 15 weeks old, shows no horns us yet (September 4 ) Bui. 1346, U. S. Dept. of Agriculture Plate V B2500M Fig. I.— View on Washoe Game Refuge, Nev. Photograph taken in the spring of 1924 on the antelope plain in northwestern Nevada. Typical antelope range, the vegetation consisting of sagebrush, with some low grasses and herbs B2729M Fig. 2.— Home of Antelope in Grand Canyon, Ariz. Hermit Creek Camp, where 11 young antelope were placed by the Biological Survey in September, 1924. Here they were fed at first by the National Park Service and are prov- ing a great attraction to tourist visitors STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 17 to concentrate and care for the young animals when captured and to raise them on the bottle. This new plan of operations was submitted to the donors of the fund and heartily approved. Its practical execution was placed in the hands of E. R. Sans of the Biological Survey. Forty newly born fawns were captured in the spring of 1924 by Mr. Sans, and all but a small number were safely reared during the summer and distributed in fall, as detailed below. The outcome of this experiment has been the placing of 12 ante- lope in Hermit Basin, in the Grand Canyon National Park, Ariz., where it is hoped they may thrive and increase so as to give pleasure to many thousands of visitors during the coming years. (PI. V, fig. 2.) Ten of the young antelope were placed on the Niobrara reserva- tion, near Valentine, Nebr., 9 on the National Bison Range, in west- ern Montana, for the purpose of attempting to build up herds of these animals on both of these Federal game refuges, and 2 in the chw park at Reno, Nev. It is obvious that, following Mr. Sans’s methods, the capture of young antelope would be perfectly simple in southeastern Oregon, in various parts of Wyoming, and in other districts where consid- erable numbers of these animals still exist, especially in areas like the Greybull River section of Wyoming, where they have increased until they are looked upon with disfavor b}^ many of the resident farmers. To perpetuate antelope under fence, even in game refuges cover- ing large areas, experience has shown that very great precautions must be taken first to destroy predatory animals, as bobcats and coyotes. Antelope within such areas appear to lose their freedom of movement and become extraordinarily helpless. This is par- ticularly the case during heavy snowstorms, when they remain within more or less definite areas, in which predatory animals capture them with surprising ease. The antelope herds in the Wind Cave Game Preserve in South Dakota and on the National Bison Range in Montana, the latter area consisting of more than 18,000 acres under fence, were brought up to a total of about 100 animals. Predatory-animal hunters had been detailed repeatedly by the Biological Survey to kill coyotes and bobcats in and about these refuges until the number of animals thus destroyed amounted to several hundred. Notwithstanding tli is, however, during severe winter storms in two seasons the band of 04 antelope on the Bison Range was completely destroyed by wandering predatory animals, which were able to drive them into snow drifts and kill them without difficulty. More than half the herd on the Wind Cave refuge also was killed, partly by coyotes and partly by bobcats. Evidence as to the danger from the bobcats was made plain when the Biological Survey warden, riding through the open pine forest of the Wind Cave refuge during a snowstorm, found and fol- lowed the fresh frail of a solitary old buck antelope. He soon came upon the tracks of a bobcat which had taken up the trail also. A short distance beyond he found the antelope, just killed and still warm. It was a full-grown buck in good condition and apparently had been easily killed by the bobcat, which had leaped upon its back. 44349° — 25 3 18 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE The difficulties which have attended the establishment of a band of antelope on the Wichita fenced game preserve are detailed else- where in this bulletin. With the reintroduction of antelope on the Bison Range from the Nevada fawns, the placing of a small band on the Niobrara Reserva- tion, and the additions to the band still existing on the Wind Cave Refuge by the Biological Survey, and those on the Wichita Game Refuge by the Forest Service, and in the Grand Canyon and in the Yellowstone National Parks by the National Park Service, the Federal Government is now attempting to build up herds of prong- horns in six widely scattered localities within the limits of their former range. METHODS OF CAPTURING AND TRANSPLANTING ANTELOPE The following statement by E. R. Sans, supervisor of predatory- animal-control work of the Biological Survey in Nevada, who suc- cessfully directed the capture and rearing of the 40 young ante- lope in northwestern Nevada during the spring of 1924, contains so interesting and straightforward an account of the methods followed that it should enable anyone to repeat the operations successfully wherever any considerable number of antelope occur : In the northern part of Washoe County, where we captured the young antelope fawns, I estimate that there were from 1,000 to 2,000 antelope ranging during the year except in the winter months. During the fall of 1923 they began leaving this range the latter part of November and began returning the first of March, 1924. During December, January, and February they ranged in the High Rock Canyon country, about 40 or 50 miles south of their summer range. In a letter received early in May, 1925, Mr. Sans sums up his latest impres- sions concerning the antelope of this section as follows : The bunch at Last Chance, where we took the fawns last year, generally leave the plateau coun- try about the last of November and work both ways, north and south, part of them going down into Guano Valley along the Oregon border and, I believe, crossing into Oregon, and the others south down the High Rock close to the Black Rock Desert. The large bunch that ranges during the summer east of Guano Valley, in the high plateau country drift the same way, some going into Guano Valley and others into Virgin Valley and down toward the Black Rock Desert. On April 18, 1924, I visited the .summer range, and while riding on horseback over one of our predatory-animal trap lines I saw antelope everywhere I looked in bunches of 3 to 9. The does were becoming heavy with fawns, and I looked for them to start dropping them about the first of May. In order to be ready when the first fawns were dropped, I selected three men to start working on May 1. They were to establish camp at the Last Chance Ranch, owned by the Hapgood brothers, located at the head of the antelope range. O. C. Wood, Leo Weilmunster, and True Hapgood, one of the owners of the ranch, made up our crew. They were to ride the range each day, watching the female antelope to learn when the first young were dropped. They were beginning to become discouraged when, on May- 17, they discovered the first new-born fawn. On May 19 I arrived at the Hapgood Ranch accompanied STATUS OF THE PEONGHOENED ANTELOPE, 1922-1024: 19 by Smith Riley, in charge of reservations, Bureau of Biological Survey, aud found the boys had 7 antelope fawns in one of the rooms of the house. They were awkward-looking things, all legs and ears. LOCATING; AND CAPTURING FAWNS Our method of capturing the fawns was to ride out on the range early in the morning, keeping ourselves from view and using powerful field glasses in looking at the scattered antelope until we discovered a female that showed she had dropped her fawns. Then it was a matter of keeping her in view until she went to feed her young, which she had hidden out, generally one in a place, possibly 75 to 100 yards apart [PI. II, fig. 1.] Along about 8.30 to 10 in the morning she would go to water, then gradually feed back toward her fawns, never looking toward them but feeding as unconcernedly as if she were merely getting her morning’s breakfast. Suddenly you would see a little speck raise up and the mother squat so that it could nurse. She would allow it to suckle a few minutes and then go on, the fawn following her until she picked up the twin. Then the same process would be repeated. The mother then would feed about possibly 20 minutes or a half hour, accompanied by the fawns, when one of them would drop down and she would lead the other about 75 to 100 yards and leave it also lying down. One particular doe that I was watching stopped suddenly and one of the fawns started running at right angles from the direction the mother was going, and when about 50 yards away dropped down as if it had been shot. The mother fed gradually on, leaving it there. After the mother had left the immediate neighborhood of the hidden fawns we took particular care to sight up with objects so we could ride to the fawns. They were generally lying in the sun on the lava rocks with no shade to protect them, their heads stretched out on the ground, ears lying flat on their heads, and very difficult to see. They would allow us to walk up almost beside them before they would make a move. Then they would jump up and develop a wonderful burst of speed for about 30 or 40 yards, when their legs would begin to tangle up and they would fall down. I was successful in capturing two of them one morning, and was in doubt as to which of us was the most exhausted by the race when we both fell at the end. As soon as the young wTere captured they were placed in a grain sack with a hole cut in the side about 4 inches from the bottom, just large enough for the head to be put out. [See PI. II, fig. 2.] One was hung on each side of the saddle horse and thus carried to camp, where they were kept until a sufficient number were caught to warrant a trip to the permanent inclosure at the ranch. (PI. Ill, fig. 1.) The mothers are very suspicious and will not go near the fawns if they can see or scent a person ; therefore much caution must be exercised in selecting a hiding place when watching to locate the young. FEEDING ON THE BOTTLE The next most important thing is food for the young. AVe found that to begin with, the best ration was rich cow’s milk, about two-thirds of a pint at a feeding, heated to a little more than the body temperature and fed from a bottle through a nipple known as a lamb’s nipple. Great care must be exercised not to overfeed. As the fawns all look alike, when you get a lot of them together it is necessary as soon as each one is fed to separate it from the others. Overfeeding will probably cause scours, which are difficult to handle. In the raising of our 40 kids we had only one that developed a real case of scours. While we brought it through by a liberal dose of lime water and 20 BULLETIN 1346, U. S. DEPARTMENT OE AGRICULTURE castor oil, it never did so well as those that had not had it. We also found that Avlienever one of them looked a little dumpy, a tablespoonful of castor oil invariably brought it around within 24 hours. The warn milk was fed three times a day, 6 o’clock in the morning, noon, and 6 o’clock in the evening. (See PI. Ill, fig. 2.) GRAIN AND GREEN FEED When the fawns were six weeks old we began to feed them a little wheat (bran and middlings) and when the;7 were two months old quite a number were eating it. Then we substituted steamed rolled oats, which they ate readily at the same time that they were also eating considerable green feed from the meadow. When they were about 2% months old the green feed dried up on account of the extremely dry season, and the fawns began to fall off in flesh, but ate about twice as much grain as they did when the green feed was plentiful. When they were 3 months old we moved them to Reno, so as to be able to get green feed, as it would have been hard on them to wean them from their milk without it. They were very fond of green lawn cuttings of blue grass and white clover, and we cut it from the lawn three and four times a day, feeding it to them fresh and keeping an ample supply of rolled oats where they could go to it at all times. They did not appear to miss the milk, which was discontinued on August 25, with the exception of five or six weaklings and cripples which were fed milk until the 5th or 6th of September. Even though we had discontinued the milk, the kids began to fatten up and grow fast upon the green-feed and rolled-oats diet. We shipped some of them on the 9th of September, when their average weight was 42 pounds, at 3% months of age. They ranged in weight from about 35 up to 50 pounds each. We purchased a bale of first-cutting, well-matured alfalfa hay, keeping a supply in the inclosure from the time they were 3 months old, and they would feed a little at a time on this, and no doubt if they did not have an ample sup- ply of green feed they would have taken to it very readily. For a winter diet I believe well-cured meadow hay and rolled oats would be the best. In feeding alfalfa hay care should be exercised not to get unmatured second or third cuttings, as there is danger of its causing bloating. Mr. Green, of the Cali- fornia Fish and Game Commission, told me that they had lost several deer and elk from allowing them to feed on green alfalfa in the field. TRANSFER TO TEMPORARY INCLOSURE When a sufficient number of fawns were caught to make a load they were placed in an automobile and taken to the Wood ranch at Diessner, Nev., where we had built a wire-fenced inclosure containing 5 acres of meadowland. The distance by road was about 55 miles, but going by saddle horse it was only 12 miles. It was much easier on the fawns to go in a car, even a much greater distance, than it would have been on pack horses for a shorter distance. In building temporary inclosures to hold antelope fawns until they are old enough to transfer to their permanent homes (PL IV) much care should be taken to have all posts on the outside of the- wire netting and all braces so placed that there will be no projections inside for the fawns to run against. The feed- ing corral should be small and made of dressed lumber with no cracks, or if net wire is used it should be lined inside with sacking to keep the animals from getting hurt when they become frightened. STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 CARE TO PREVENT FRIGHTENING 21 Antelope fawns are very panicky when something unusual occurs. The attendants must move slowly, talking to them at all times; quick movements frighten them. Upon entering the place where they are confined, if one will talk to them and be careful to move slowly, in a few minutes they will settle down and allow him to work among them without signs of fear. Even after they are 2 or 3 months old it is necessary to use much caution when going among them, but I noticed that the fawns usually come running to a man when they become frightened, apparently for protection. When suddenly frightened, however, they may rush at full speed into a fence without appar- ently seeing it. With a single exception every fawn we lost was injured or killed outright from being frightened. Two of them broke their necks by hit- ting the wire fence on a slant, running their noses into the mesh. Fawns apparently have a hereditary fear of cats and dogs, and I believe cats scare them worse than the dogs. I have been wondering if the bobcat family isn’t one of their worst enemies on the range. We found out when we crated the animals at Diessner that one must be very careful, for after a few were caught the others became frightened. At first we caught all the gentle ones, leaving the wild ones until the last. When we came to the last two, which were the wildest of all, one of them made a frantic jump at the wire fence, striking it on the slant and' breaking its neck. When we recrated them at Reno I purchased lumber and made a corral about 8 by 20 feet with a lumber chute narrowing to about 4 feet wide at one end. Putting all the fawns into this small inclosure, we picked the wildest one first and by catching the four or five that did not come to us readily when they were out in the big inclosure we had no trouble ’whatever. The old gentle stand-bys did not become frightened, and within a few minutes we had the total number for shipment crated. Fig. 2. — Specially constructed crate for shipping 4-months-old antelope. Made of five-eighths-incli material and requir- ing 20% feet of common lumber, the inside measurements being 38% inches long, 33 inches high, and 13 inches wide. Weight, 35 pounds. Cost of making crate and lining with burlap, including labor and materials, $2.40 SHIPPING CRATES Crates should not be built too large. (Fig. 2.) A crate for an antelope averaging 50 pounds in weight should be about 38% inches long, 13 inches wide, and 33 inches high, inside measurements, using select dressed lumber, about one-half or five-eighths of an inch thick by 3 inches wide. The bottom and sides up for about 12 or 14 inches should be solid, and then 3-inch strips placed 3 inches apart on the sides and ends to allow good ventilation. The two strips above the solid bottom board should be wrapped with burlap so as to make a padded crate. It is well to take a small joining plane and run over the edges of the boards so that there will be no sharp edges to cut. The main thing in making a crate is to have it so narrow that the animal can not change ends in il. Antelope should be put in the crate backward with head toward the door. The door should la; made so that it will slip in a slot 22 BULLETIN 1346, U. S. DEPARTMENT OP AGRICULTURE at the bottom and then can be tied at the top with stout cord. Room for a water can and space for grain should be allowed in the front. Coarse straw at least 3 inches deep should be placed inside for a bed while in transit. The express company charges one and one-half times the first-class rate on live animals, but does not include in the weight the feed necessary while en route. Water and feed should not be placed in the crates, therefore, until after they are weighed and billed out. TRANSPORTING ANTELOPE IN CRATES Animals should be crated singly, for where two or more are together they will cut each other with their sharp hoofs. When one lies down the other will invariably step on it. In the 240 miles that we transported the fawns from Diessner to Reno by automobile truck they were crated singly and endured the trip in fine shape. We left Diessner at 2 o’clock in the afternoon and, driving all night, arrived in Reno at 8.15 in the morning, thus making the trip during the coolest period of the day. We have demonstrated beyond a doubt that young antelope can be success- fully taken from their natural range and raised if care is exercised in handling them. Great care must be taken to keep all milk containers, bottles, and nipples perfectly clean, and I would suggest that if any more antelope are taken in this district they be transferred direct to Reno as soon as the required number is captured, where unlimited supplies of milk can be obtained and also all other necessary food. It is not necessary to have such a large inclosure as we made at the Wood ranch, and I believe the antelope become gentle more readily where they see people constantly passing. RESULTS OF A CENSUS OF EXISTING ANTELOPE For many years the Biological Survey has been engaged in deter- mining the former and present range of the pronghorn. In 1922 it became evident that the time had arrived for active measures leading to their conservation if their extermination was to be avoided. In order to form a basis on which intelligent conservation measures might be built a definite census of the surviving pronghorns was undertaken. This was continued from 1922 well into 1924 through field men of the Biological Survey, with the active and friendly co- operation of State game officials, State game protective associations, and individuals, not only in the United States but in Canada and Mexico, involving a great deal of correspondence. In taking the census of the antelope in the United States the sur- vey has been fortunate in having a field organization for the control of predatory animals and of harmful rodents in each of the 16 States where the pronghorn still occurs. Their operations in these services are of State-wide character and are conducted in cooperation with the State extension services, State departments of agriculture, and other organizations, as well as stockmen and farmers. Their work puts them in touch with county agents, sheriffs, and other officials, hunters, and men generally familiar with the State. The State game officials were particularly helpful, and the field force of the United -States Forest Service also added many facts. Without the contributions from these varied sources this report could not have been prepared. Each contributor is entitled to feel that he has STATUS OF THE PRONGHORNED ANTELOPE 1922-1924 23 definitely assisted in a work which may be the basis for definite action in various States to insure the perpetuation of the pronghorn. In taking this census determined efforts were made to locate as completely as possible each surviving band and, wherever possible, to have an actual count of the animals in it. Almost throughout its range the pronghorn is decreasing. Each succeeding year some of the smaller herds marked on the accom- panying maps are certain to disappear, and only in the most favor- able* areas, where they are carefully protected, is there hope for the long survival of these interesting animals. In perpetuating herds of antelope in the different States one of the principal factors will be the interest taken in them by ranchmen, local sportsmen, and other residents. Antelope are on the verge of final extermination in Kansas, where in the early days they were familiar sights from the windows of passing trains. There is little hope for the preservation of the large number of small bands containing from three to a dozen or more pronghorns. Under present conditions, when a band is reduced to a very small number, its continued existence is practically impossible unless it has the benefit of exceedingly careful guardianship. The decrease of antelope is governed by a number of conditions, among which may be mentioned the inroads of predatory animals, illegal shooting, and the increased occupation of their territory for economic uses and the disturbance brought about by it. There may be improvement as to the first two of these factors, but the last is one which is necessarily beyond control. This means that eventually the surviving antelope will be limited to bands located in some of the more desert and least occupied parts of their former range, such as in northwestern Nevada, or to large, fenced game refuges. There are areas in many of the Western States which are suitable for the maintenance of bands of antelope on the open range if public sentiment will interest itself in them. The largest herds of antelope in any restricted area appear to be located on the great plains which cover northwestern Nevada and adjacent parts of Oregon and southwestern Idaho. In this iso- lated, sparsely populated region may still be found the nearest approach to original conditions of any part of the antelope range in the United States to-day. There is no intention to claim anything like complete accuracy in all the specific localities and numbers of antelope set forth. From the very nature of the case that is obviously impossible. Furthermore, with the decrease of antelope some of the small bands here listed may have already disappeared. Also occasional bands here and there may have been missed, and the numbers estimated for many may be erroneous — either too great or too small. It is earnestly desired that all who are in position to give constructive criticism will send in the information which they possess in order that corrections and additions may be made on the records. Table 1, on page 3, summarizes the number of bands of antelope and the total number of animals in each within the various geo- graphic areas. Following is a detailed statement showing the loca- tion and distribution of the bands in each of the areas listed. 24 BULLETIN 1346, U. S. DEPARTMENT OE AGRICULTURE ARIZONA In 1923 antelope in Arizona were restricted to bands occurring in 18 areas and totaled about 650 animals. A more intensive investigation might slightly increase this number. Formerly they existed in great numbers in this State, where the range and climatic conditions were peculiarly favorable to them. In the eastern part of the State from 1SS4 to 1S89 they were very numerous both on the grassy plains and in the parklike openings among the scattered cedar, juniper, and pinyon forests covering vast areas of that great plateau region. In summer they ranged also among the yellow-pine forests on the Mogollon Plateau, but the heavy snows on these higher elevations forced them into the lower country in winter. As elsewhere at that period, there was no appreciation among sportsmen and the general public of the need of any re- straint in killing game, and both antelope and other kinds of big game were killed freely throughout the year. With the increasing occupation of the ranges the antelope have steadily decreased and now are extinct over great areas where they once abounded. The information concerning the number of antelope now in Arizona has come mainly from M. E. Musgrave, in charge of the predatory-animal work, and D. A. Gilchrist, of the rodent-control work, both of the Biological Survey. Their personal familiarity with all parts of the State has enabled them to make an excellent preliminary survey of the situation. On December 15, 1923, Mr. Musgrave wrote : “ We have collected fairly accurate data on the antelope of this State and I have found that the number of young this year ranges from 10 to 25 per cent of the total number of animals a year or more old. The heaviest percentages of young are on the ranges in the vicinity of Valentine, north of Seligman. It is believed that there were 25 per cent of young born in the Sitgreaves Na- tional Forest and only about 10 per cent on the Prescott National Forest and adjacent to it. The small band located in the San Fernando Valley, south of Tucson, had an increase of about 20 per cent. The total increase for the State for the present year should be about 15 per cent. “ The antelope that live on the Anderson Mesa have been materially reduced in number. Last year they numbered more than 100 head, but now only 25 survivors are reported. I am inclined to believe that the antelope in Arizona are increasing regardless of the rapid depletion of the herds on the Anderson Mesa, south of Flagstaff, and along the Verde rim. Antelope appear to be increasing on cattle ranges in the State and decreasing on sheep ranges.” The distribution of antelope in Arizona is approximately as follows (fig. 3) : 3 1. About 12 antelope are reported as ranging in Antelope Valley, between Hurricane Ledge and Kanab Creek, north of the Grand Canyon, in Mohave County. 2. In 1923, according to residents of Grand Canyon and railroad men em- ployed on trains between Williams and Grand Canyon station, in Coconino County, about a dozen antelope still occurred in the vicinity of Anita, or be- tween Anita and Cataract Canyon. They are seen occasionally from the trains. 3. About 50 in small scattered bands occur in open parks in the yellow-pine forest north of Flagstaff, in Coconino County. 4. Supervisor Miller, of the Coconino National Forest, reported in 1923 that the antelope ranging on Anderson Mesa, southeast of Flagstaff, in Coconino County, in 1913 numbered about 200, but that in 1922 they had become reduced to about 100, and in 1923 to about 25. This decrease indicates the danger of extermination confronting this herd. 3 The paragraph numbers in this and other States and Provinces correspond to the numbered, areas shown on the respective distribution maps, the number of the area being the figure outside the circle, the number of antelope in each area being expressed by the figure within the circle. Bui. 1346, U. S. Dept, of Agriculture Plate VI Antelope in Wind Cave Refuge, S. Dak. As shown in the illustration, antelope sometimes eat snow and then arc able to get along in winter without water STATUS OF THE PBONGHOKNED ANTELOPE, 1922-1924 25 5. Formerly many antelope ranged on the sagebrush plains southwest of Winslow, in Navajo and Coconino Counties, but in 1923 these had become re- duced to 10, located in Coconino County. 6. About 50 are reported on the open range near Heber, in Navajo County. 7. About 50 are reported as ranging along the Verde rim, northeast of Camp Verde, Yavapai County. S. About 40 are ranging in Cedar Glade, south of Ashfork, in Yavapai County. 9. In 1922 several bands, aggregating 100 or more, were reported to be rang- ing on the Baca Grant, 50 miles south of Seligman, in Yavapai County. 10. A band of about 50 is reported as ranging in the open country near Selig- man, in Yavapai and Coconino Counties. 11. In 1922 about SO were reported to have been on the Carrow cattle range, southwest of Nelson, in Mohave County. These probably also range into Yavapai County. The Car- row brothers' give strict protection to the antelope on their range, where they have increased within the last 10 years from about 15 animals to the present number. 12. A band of 25 is re- ported as occupying the open range southeast of Springerville, in Apache County. From 18S4 to 1890 bands aggregating 500 to 600 antelope occupied this territory. 13. In 1923 bands aggre- grating about 75 antelope were reported to have been ranging in Yuma County, n ear the international 1 toundary. Seven head were reported in 1924 between the Mohawk and the Ca- beza Prieta ranges. 14. In 1923 Papago In- dians reported that a few antelope were still ranging in Santa Rosa Valley, in Pima County, but no defi- nite number was given. Fig. 3. — Distribution of antelope in Arizona; estimated 15. In 1923 about 30 ante- at 651, in 18 area's lope were 'reported occur- ring on the mesa west of Oracle and along the road to Florence, about 35’ mileS northwest of Tucson, in Pinal and Pima Counties. 16. According to residents of Arivaca, about 20 antelope occur in the upper end of Altar Valley, not far from San Fernando Valley, in Pima County. Near a small lake near the middle of the valley in 1923 a number of young were noticed with this band, indicating that it may be slowly increasing. 17. In 1923 a band of about 12 was ranging on the plains near the north end of a small mountain range locally known as the Sierrita, southwest of Tucson, in Pima County. 18. In 1923 about 10 antelope were ranging on the plains near Benson, in Cochise County. CALIFORNIA During the middle of the last century when the gold rush took place, antelope were generally distributed and very abundant on the plains of California, especially’ in (he San Joaquin Valley and over the Mohave Desert region, ranging thence south to the Mexican border; also on the arid plains in the northeastern part of the State. In 1923 they had been reduced to small band in six widely separated areas, containing a total of about 1,057 animals. 44349°— 2 JJ- 4 26 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE Although the intensive occupation of California is making nearly all parts of the State impossible for the presence of antelope in a state of freedom, there are a few areas within which herds might be perpetuated, especially in the northeast. Fortunately, the State fish and game commission is taking an active interest in the matter, as is also the committee for the conservation of wild life of the California Academy of Sciences. At the Washington Antelope Conference F. M. Newbert, president of the hoard of fish and game commissioners, stated that for a long time the game commission has taken an active part in trying to safeguard the surviving antelope. At one time the Mount Dome herd (No. 1 on the map) had dwindled to 11 animals. Through the interest of the commission in safeguarding these animals by its warden service the number has been increased until now there are about 118. Mr. Newbert expressed a desire to cooperate with the Biological Survey and also with the authorities of the States of Nevada and Oregon in any practical way to perpetuate and if possible increase the antelope herds. It will be difficult to perpetuate the four small herds in the southern part of the State, but in the northeast there is greater promise of success. A correspondent who visited northeastern California in the fall of 1922 stated that three-fourths of the dry farms have been deserted in that region, which naturally tends to restore favorable conditions for the antelope. The committee for the conservation of wild life of the California Academy of Sciences has raised funds for feeding the Mount Dome antelope in severe winters and has joined with the State game commission in having game wardens protect them from poachers. It is planned to try to have an antelope refuge established for their benefit. M. Hall McAllister, chairman of that committee, wrote on July 16, 1924 : “ By reason of the remarkably open winter of 1923-24 the Mount Dome herd remained scattered and did not band up as they usually have done : therefore, no tally was possible on them as has been usual. Also by reason of the small rainfall and few heavy storms in California the Lassen herds [No. 2 on the map] have migrated across the line into Nevada, and in July, 1924, not over a dozen were sighted on their old grounds in Lassen County, although some 500 are reported ranging in the neighboring part of northwestern Nevada.” No doubt the shifting of the Lassen County herds to the adjacent part of Nevada was caused by shortage of feed in the California area, and these animals may be expected to return to Lassen County as soon afe adequate rainfall again restores the proper forage conditions. For this reason this herd is being credited to California, which appears to be their natural home. Fortu- nately, the area they have gone to in Nevada lies within the antelope refuge established in 1923 by the governor, thus affording them protection. In California, as elsewhere, coyotes and other wild animals prey upon the diminishing herds of antelope, as upon other wild life. Wherever it is prac- ticable the Biological Survey is having its field leaders give special attention to the destruction of predatory animals endangering the surviving herds of antelope and other large game. Hunter Fay Clark, working cooperatively for the Biological Survey and the California Department of Agriculture in the Mount Dome district, has been successful in destroying coyotes which have been preying on the antelope and deer of this district. The inroads of these pests undoubtedly account for the slow increase of this herd. In a single month, in the range of these antelope, Mr. Clark killed 35 coyotes. The result of his work will undoubtedly become apparent in an increase of the game animals of that district, as has been the case following similar work in other parts of the West. STATUS OF THE PEONGHOBNED ANTELOPE, 1922-1924 27 The detailed information concerning the distribution and number of antelope now in California has been supplied mainly by M. Hall McAllister, of the California Academy of Sciences ; George Neale, executive officer of the Cali- fornia Fish and Game Commission ; and F. E. Garlough, of the rodent-control section of the Biological Survey. The distribution of antelope in California is approximately as follows (fig. 4) : 1. John O. Miller reported on February 2S, 1923, that the Mount Dome herd contained about 118 antelope. They range on the plains in the southern end of Lower Klamath Lake, near Mount Dome, in Siskiyou County, probably the most favorable area in California for perpetuating antelope under natural conditions. For some years the herd has had the attention of the State fish and game commission and other conservationists, as set forth above, and with the destruction of preda- tory animals in that dis- trict should increase in numbers. 2. The Lassen County antelope, scattered in nu- merous small bands on Madeline Plains, in Dixie Talley, Secret Talley, and other points, range east- ward into the Smoke Creek Desert of Nevada. These were counted by W. J. Lee, who reported on September 29, 1923, that they aggregated S64' ani- mals. Since then it has been reported that most of these animals have crossed the State boundary into the adjacent part of Nevada, as a result of fail- ing forage due to the ex- cessively dry season of 1924. They will undoubt- edly return with the res- toration of more favorable conditions. 3. In 1922 two small hands, totaling 29 animals, were reported as ranging between Mendota and Pa- noche Creek, on the west side of the San Joaquin Talley, in Fresno County. 4. A hand of about 30 was located in 1922 between Granite Wells and Rands- hurg, on the Mohave Desert, in San Bernardino County. H. In 1922 a band of 11 lived in Antelope Talley on the Kern-Los Angeles refuge, ranging over adjacent parts of Kern and Los Angeles Counties. In April. 1924, 13 were reported as having been seen between Willow Springs and Liebre Ranch, on the west side of the valley in Kern County. 0. A band of 5 was reported in 1922 ranging in the desert north of the State highway between Campo and Imperial, in Imperial and San Diego Counties. These animals probably range back and forth across the Mexican border. COLORADO The plains of Colorado formerly abounded in antelope. It is surprising that with the long-continued occupation of this State by farmers and stock- men bands of antelope still survived up to October, 1923, in 28 localities, ag- gregating approximately 1,233 animals. This indicates an interest in or at Fig. 4. — Distribution of antelope in California, esti- mated at 1,057, in six areas 28 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE least a tolerance of these animals, which might well be converted into such an active spirit of protection as would perpetuate a number of herds in suitable localities. The very excellent census of antelope in Colorado here presented is due largely to the persistent efforts of John H. Hatton, secretary-treasurer of the Colorado Game and Fish Protective Association, and to the work of Stanley P. Young and Joseph Keyes, in charge of the predatory-animal and rodent- control work of the Biological Survey in the State. Mr. Hatton wrote that the active interest of the Colorado Game and Fish Protective Association in the conservation of antelope will undoubtedly he very helpful in bringing about their perpetuation. Early in 1923 the legislature of Colorado established the Colorado Ante- lope Refuge in northern Larimer County, on the Wyoming line, which covers the territory occupied by a band of about 75 animals. The fall of the same year Mr. Hatton wrote concerning the plans of the Colorado Game and Fish Protective Association : “We plan during the next legislature to have presented some areas in the eastern part of the State which would he suitable for antelope refuges. We first became interested in this subject as an organization a couple of years ago, and it seemed that the first logical step would be to get a census of the species, which, as you know, has already been done. I plan also a little later to address' a letter to the local newspapers where antelope are found, giving a little account of them and making appeals for their pro- tection.” The general results of inquiries concerning antelope in Colorado indicate that these animals are steadily decreasing, especially on the eastern plains. There has been some discussion of the possibility of rounding up the sur- vivors in this region and placing them under fence in northeastern Larimer County in the antelope refuge recently established. Antelope once abounded in North Park, but the county assessor of Jack- son County, in a letter dated October 31, 1923, stated that none remain and that the last band, containing 21, was killed many years ago by hunters from Fort Collins. About 35 years ago in that district a hunter used a, telescope rifle for killing antelope, for which he was paid $1 each. Wagons were sent out to pick up the animals, which were shipped to Denver for sale. The results of the census of antelope in Colorado indicate that there is still time to build up fine herds of these beautiful animals in parts of the State where they will not become obnoxious to the farmers. A careful study of the situation should be made for the purpose of locating satisfactory areas where definite efforts can be made to have State refuges established, and in addition of enlisting the cooperation of men having large fenced pastures to maintain small antelope herds as a matter of interest. Apparently one of the vital needs in Colorado, as in many other States, is to insure a better enforcement of the law against those who wantonly kill antelope. A large proportion of these animals now being killed are, no doubt, shot merely for passing amusement. The distribution of antelope in Colorado is approximately as follows (fig. 5) : 1. Not less than 150 antelope range between Vermilion Creek and the head of Sand Wash in northwestern Moffat County. The majority of the people in that district seriously object to these animals being killed, although it is stated that occasionally a resident shoots one for his own use. 2. In Middle Park, northwestern Grand County, a small band of 7 still exists. This was formerly a favorite range for antelope, but they have been rapidly decreasing there and will probably be completely exterminated in the not distant future. STATUS OP THE PRONGHORNED ANTELOPE, 1922-1924 29 3. A band of about 75 is located in northeastern Larimer County, on the Colorado State Antelope Refuge, mentioned above. These antelope, no doubt, range across the boundary into southern Wyoming, and are said to be holding their own and possibly increasing. Some of the people in this district favor the protection of the antelope, while others do not and are reported to look upon them as a nuisance. 4. About 33 antelope are reported from north-central Morgan County, where they are said to be decreasing. 5. In Logan County about 150 survive. There is some complaint here of their destroying crops. Although antelope appear to be occasionally shot in this district, the herd appears to be holding its own and with a little better protection would undoubtedly increase. It is obvious that the relation of these antelope to the farming community requires careful study before any definite effort is made to build up the herd. 6. C. F. Parker wrote from Julesburg in October, 1923, that about 55 ante- lope are located in the northwestern corner of Sedgwick County, where the farmers desire their protection. In winter they come down and feed in the Fig. 5. — Distribution of antelope in Colorado, estimated at 1,233, in 28 areas. Antelope refuge indicated by broken lines alfalfa fields. They are haid to be increasing. Mr. Parker stated that when he fenced an 8,000-acre pasture at his Cottonwood Ranch about 20 years ago 2 does and 1 buck antelope were included. They have steadily increased since that time, and it is generally understood in the neighborhood that any- one caught molesting them will receive the limit of the law. Occasionally in that district antelope are pursued and shot by hunters in automobiles, but the resident farmers are interested in them and desire their protection. 7. A small band of eight has about held its own for several years in south- eastern Phillips County. The people in that district are interested in them and favor their increase, which no doubt could be brought about by a little more careful protection. 8. A band of 13 is located in northeastern Yuma County. 0. A band of 22 is reported from northwestern Yuma County. 10. In northern. Washington County, the existing herd numbers about 48, and their protection is favored by most of the people in the county, although occasionally one is shot. Suggestions have been made that they be captured and removed to a fenced refuge, hut this is not generally favored. 30 BULLETIN- 1346, U. S. DEPARTMENT OP AGRICULTURE 11. A band of about 15 lias been reported from southwestern Washington County. It is stated that both this and the band in the northern part of this county are about holding their own. H. It. Rice, assessor of Washington County, wrote under date of October 27, 1923, that about two months previously 2 antelope had been -shot north of Plattner and left as they fell, and that 3 Avere killed in the same neighborhood a year or so before, but that efforts to locate the killers were unavailing. Mr. Rice thinks that the partial protec- tion the antelope receive has a tendency to cause them to lose their fear of man, and this enables poachers to kill them with little difficulty. He thinks that they could best be perpetuated by inclosing them in a pasture from which each year the surplus bucks might be killed by hunters. This idea appeared to appeal to some of the local sportsmen. 12, 13, and 14. Three bands of about 20 each are reported from Adams County. It is stated that there is not much local interest in them and the herds are decreasing. 15 and 18. About 27 antelope are said to remain in the western part of Elbert County and 150 in the northeastern part. Their protection is favored by most of the people living in that district ; but it is reported that they are frequently hunted, especially by men in automobiles. NotAvithstanding this, the herds are reported to be about holding their ovrn. It has been suggested that an antelope refuge might be established for the larger of these two herds. 17. A band of 10 or more antelope is reported in east-central Lincoln County. There is some not very active sentiment for their protection, and through killing by poachers the herd is decreasing. 18. In northwestern Cheyenne County about 50 antelope still remain. There is said to be little sentiment in favor of their protection, and they are being killed by hunters. 19. About 45 antelope occur about 10 miles north of Buena Vista, on the line between Chaffee and Park Counties. Local sentiment is very favorable to their protection. 20. In southAvestern Park and eastern Chaffee Counties is a band of 40. The majority of the residents in this section favor their protection, but the band is decreasing, probably through hunting. 21. In July, 1924, a band of about 40 antelope was reported in east-central El Paso County. 22. In southeastern Fremont County a band of about 10 still survives. Their protection is favored by residents, but occasionally one is killed by hunters. 23 and 24. In July, 1924, a band of 10 was reported in the east-central part of Pueblo County, and one of about 150 in the south-central part. 25. About 40 are reported from near Crestone, in southeastern Sagauche County, on the Luis Maria Baca grant No. 4. They are under fence in this grant and as a result of the protection they are receiving are reported to be increasing. 26. A band of 7 is reported in the extreme eastern part of Huerfano County. 27. A band of 10 ranges in north-central Las Animas County. 28. A band of 8 is reported in the northwestern Baca County. IDAHO Bands of antelope are reported in 14 areas in Idaho with an aggregate of approximately 1,500 animals. They are located mainly in the east-central and extreme southwestern parts of the State. Information concerning them has been obtained from R. E. Thomas, State game warden ; United States Forest Superintendent Olsen ; and L. J. Goldman, leader of the Biological Survey’s predatory-animal work in the State. During the past four years considerable work has been done to bring about the establishment of a Federal antelope and sage-hen refuge, covering the Owyhee desert country in the extreme southwestern part of the State, in Owyhee County. Details concerning this are given elsewhere in this bulletin. The present distribution of antelope in the State is approximately as follows (fig. 6) : 1. A band ranges in Antelope Valley and the Pahsimeroi Mountains in Cus- ter County. Forest Supervisor Olsen states that they inhabit the low open ridges lying at the base of the mountains generally, but occasionally a buck 31 STATUS OF THE PRONG-HORNED ANTELOPE, 1922-1924 is seen in tlie higher elevations, even up to 8,000 feet. Mr. Olsen states that at one time he counted 104 antelope in this band. 2. Forest Supervisor Olsen states that three years ago a band of 12 adult antelope made the Sulphur Creek ranger station in Pahsimeroi Valley, Lemhi County, their headquarters. They became very fond of alfalfa, which grows plentifully there, and have remained there ever since, except that they some- times go back into the hills a couple of miles or so during the winter season. They have been carefully protected by the forest rangers and have increased until the band now numbers 62. Forest Examiner S. B. Locke writes that at this time antelope in the Pahsimeroi Valley have become a nuisance at several ranches. At the ranch just about the ranger sta- tion 20 to 40 antelope enter the fields during the summer and consume ap- proximately one crop of alfalfa. While the hay is tall they do not cause much damage but feed on it intensively soon after it has been cut. Some of the ranchers keep them away from their fields by the use of dogs and shot- guns. Any increase in the numbers here would in- tensify the losses of the farmers. 3. A band of 9 is lo- cated near Goldburg, in Custer County. 4. A band of 25 is re- ported in Lemhi Valley, in southern Lemhi County. 5. About 25 live on Medicine Lodge Creek, in Clark County. 6. About 60 range on Birch Creek, in western Clark County and adjacent parts of Jefferson and Butte Counties. 7. About 75 are reported as ranging in Little Lost River Valley, in Butte and Custer Counties. Antelope from this area are said to range sometimes far out on the Snake River desert. 8. In July, 1924, a band Of 4 was seen on the Snake Fig. G. — Distribution of antelope in Idaho, estimated at River desert at Arco, in 1,485, in 14 areas Butte County. 9. A band of 26 was counted in 1923 in the Copper Basin, near Mackay, in southern Custer County. 10. A band of 7 exists near Chilly, on Big Lost River, in Custer County. 11. About 13 are reported to range about Horse Pleaven Pass, at the head of Pahsimeroi Valley, in Custer County. A long-time resident of that section states that a few years ago this herd contained about 100 animals. 12. A band of 25 is reported as ranging on Succor Creek, in Owyhee County. 13. Southwestern Owyhee County is the most important area in Idaho for antelope. Reports state that two separate bands occur, one ranging from the Juniper Mountains in Owyhee County to the Nevada line, and the other occupying the country from the crest of the same mountains westward to the confluence of the Owyhee River and Soldier Creek in Oregon. L. .1. Goldman, in charge of the predatory-animal work of the Biological Survey in Tdaho, wrote that he had reports from authentic sources of from 600 to 1,000 occupy- £2 BULLETIN 1346, IT. S. DEPARTMENT OF AGRICULTURE ing the extreme southwestern corner of Owyhee County. They range from the Duck Valley Indian Reservation west to the Oregon line and probably into Jordan Valley, Oreg., and from the Nevada State line to a point about 30 miles north. Stragglers and small bands undoubtedly stray beyond these limits. They also cross southward into Nevada. Their main summer range is about the forks of the Owyhee River and the Juniper Basin. E. Grandjean, of the Forest' Service, wrote that this band occupies the high plateau drained by the Owyhee River at altitudes varying from 4,500 to 6,000 feet. This area is fairly well watered and overgrown with grasses and sagebrush. In the middle of it are located the low, hilly Juniper Mountains, which are very rocky and cover an area approximately 10 miles wide by 20 miles long. This main plateau, except the wooded part, is used by antelope as spring, fall, and winter range. The animals usually appear there early in April and remain until early in winter, when the snow compels them to leave for their winter range, generally believed to be the low desert plateau lying south of the main Owyhee River. Pig. 7. — The only band of antelope in Kansas occurs in the extreme southwestern corner ; estimated to contain S animals 14. Scattered bands numbering about 50 are reported to live on Browns Bench, along the Nevada line, in Twin Falls County. These undoubtedly range back and forth across the State line. KANSAS The only antelope definitely reported as existing in Kansas in 1923 was a band estimated to contain about 8 in the extreme southwestern part of the State, in Morton County. According to State Game Warden J. B. Doze they are reported to be more often in Oklahoma than in Kansas, passing back and forth across the line (fig. 7). At one time Kansas was inhabited by myriads of pronghorns, and for years after the construction of the transcontinental railroads they were a familiar sight to passengers on the trains. In 1923, however, they had become almost exterminated throughout the State. In a letter dated July 2, 1924, Hal G. Evarts, of Hutchinson, wrote that he had recently received reliable information that in 1916 a herd of 62 pronghorns was ranging about 25 miles northwest of Cimarron, in the Pawnee Creek STATUS OP THE PRONGHORNED ANTELOPE, 1922-1924 3 8 breaks. In 1918 lie saw seven, which wintered within a mile of the town of Cimarron. In 1921 a band of 16 spent the summer and winter about 15 miles south and west of Garden City. He has uot seen them since that time, but is of the opinion that they may still be ranging in the sandhills of that vicinity. * MONTANA In the early days the great plains of Montana contained countless thousands of antelope. The present census records surviving herds in 44 districts, mainly in the eastern and central parts of the State, with a total of approxi- mately 3,090 animals. As in many other States, the antelope situation here is precarious and needs prompt attention if the herds are to be perpetuated. The information given below is mainly the result of inquiries made by Thomas N. Marlowe, chairman of the State fish and game commission ; C. A. Jakways, State game warden; by employees of the Forest Service; and especially by O. E. Stephl, R. E. Bateman, and other employees of the Biological Survey. It should be noted that several antelope herds drift back and forth between northern Montana and Canada, particularly from northern Valley and Hill Counties. wIn severe winters, antelope frequently drift down the Yellowstone Valley into the State from the Yellowstone National Park, Wyo. Some ranchmen complain of injury to crops by antelope, one complaint relating to damage to alfalfa in Powder River County. Abandonment of ranches by a large number of dry-land farmers has restored more favorable conditions for antelope over considerable areas. Many new settlers are inter- ested in protecting the antelope, but from curiosity frequently kill a single animal to get an opportunity to taste its flesh. A considerable number of the older ranch owners who were in the State during the days when the antelope was one of the common game animals, are now taking an active interest trying to prevent the extermination of the species. In some cases they afford ante- lope the same protection against hunting on their ranges that they give their cattle. The antelope have learned these sanctuaries, and when shot at else- where immediately run to them for safety. Local sportsmen’s associations are in position to be very helpful in developing antelope conservation in the State. Thomas N. Marlowe, chairman of the State Fish and Game Commission, wrote as follows : “ The matter of further protection and propagation of the antelope is, to my mind, a very difficult one. In spite of what we have tried to do in this State, they seem to be, as stated in your letter, on the decrease and practically threatened with extinction. The only possible solution of the matter appears to be the creation of an antelope preserve somewhere in eastern Montana. This should be as thoroughly fenced as the buffalo range near here and the predatory animals destroyed. “ I believe also that a new herd should be started on the buffalo range. [It will be noted below that in September, 1924, a small herd was reestablished on the buffalo range by the Biological Survey.] I believe something can be accomplished along these lines, and if a reserve is created in eastern Montana possibly our department might be able to do .something toward financing it, if not too expensive. I am with you. in the hope that some solution can be found at the antelope conference to help remedy the situation. “ I am satisfied after having been a member of the State fish and game com- mission for more than five years that the greatest problem confronting us in this State in the protection of game is the control of predatory animals. If we could exterminate them there would be plenty of game in the State for all 34 BULLETIN 1346, U. S. DEPARTMENT OP AGRICULTURE demands, as there is no doubt tliat predatory animals destroy more than all the hunters put together. Two years ago we framed a law setting aside 25 cents from each hunting and fishing license to cooperate with the Biological Survey in the destruction of predatory animals in connection with the State Livestock Association, but I am frank to say that the amount we are expending in this respect scarcely more than kills oft the increase.” The antelope reported from Montana are distributed approximately as follows (fig. S) : 1. A band of eight is reported in the vicinity of the Sweet Grass Hills, in Liberty County. Antelope are reported to have decreased in this district. 2. Another band of eight ranges along the Marias River, in Liberty County. It is reported that a few years ago several small herds were ranging 20 to 30 miles south of Chester, on the main line of the Great Northern Railroad, in Liberty County, but since then information has been conflicting, and nothing definite has been learned to indicate whether they are still there. 3. A band of 15 antelope south of Galata is said to contain the only antelope remaining in Toole County. Pig. 8. — Distribution of antelope in Montana ; estimated at 3,027, in 44 areas 4. About 12 antelope are still ranging on the Marias River and its tribu- taries, in Pondera County. Two years ago a band of about 12 ranged on the Marias River 30 miles northeast of Conrad, but people familiar with the district state that they have now disappeared. It is reported that bootleggers in high-powered automobiles passing through this district to Great Falls make a practice of pursuing and killing antelope on the open plains. The residents resent this killing but have been powerless to prevent it. It is reported also that the Conrad and Brady rod-and-gun clubs are interested in the preservation of antelope and will assist in their protection in any way that is practicable. 5. About 20 antelope are reported to range on the Teton River, in Pondera and Chouteau Counties, about 20 miles east of Brady. 6. A band of about 10 antelope is reported about 12 miles southwest of Chouteau on the north side of the Sun River, in Teton County. This band has decreased during the past few years. 7. About 40 antelope are reported to range between Chouteau, Teton County, and Great Falls, in Cascade County. 8. Two bands, numbering about 35, range north of Great Falls, in Cascade and Chouteau Counties. 9. A band of about six is reported to be ranging in the Dry Creek country, about 10 miles southeast of Augusta, in Lewis and Clark County. STATUS OF THE PRONG-HORNED ANTELOPE, 1922-1924 35 10. About 100 are reported north of Fort Benton, in Chouteau County. 11. A band of 19 was reported on February 9, 1924, to range immediately west and north of the town of Montague, Chouteau County, sometimes within half a mile. A few years ago this band was double its present size. Another band of about 60, near the town of Square Butte, is reported to have increased about 10 per cent in the past two years. 12. Three small bands, aggregating about 20, are located near Winifred, in .northern Fergus County. One is 10 miles northeast of the town, another 18 miles, and the third is ranging near Armells Creek. The total number of ante- lope here is slowly decreasing. 13. A band of 30 was reported in 1922 south of the Little Rockies, in south- western Phillips County. 14. On January 2, 1924, several bands, totaling about 200 antelope, were reported on the range from 20 to 30 miles southwest of Glasgow, in Valley County. A few range around the head of Duck Creek, Brazil, and Dry Runs. The larger bands occupy the country that divides the heads of Little, Beaver, and Lone Tree Creeks and on down the east slope to Willow Creek. The antelope here are reported to be decreasing rapidly and likely to be extermi- nated unless better protection can be given them. It is reported that antelope occasionally cross into the northern part of Valley County from Canada. 15. A band of about 50 ranges in the Missouri breaks in northwestern Garfield County, and a smaller band of S in the adjacent southeastern part of Phillips Qounty between the Little Larb Hills and the Missouri River. 16. About 100 are reported on the Snow Creek Game Preserve in Garfield County. 17. In Garfield County a band of 70 is reported along the Missouri River in Townships 23 and 24, and a band of 30 north and east of Haxby. 18. There are several bands in eastern Garfield County, aggregating about 92 animals, of which 6S occur along Woody Creek, in Townships 40, 41, and 42, and adjacent areas, and about 24 along Big Dry Creek east of Jordan. Several years ago these and other bands in this county were decreasing, but since many dry farmers have left, they are beginning to increase. 19. A band of 60 ranges along Lodgepole Creek from Dilo to the Musselshell River in Garfield County, and another band of 30 occurs farther down in the southwestern corner of the same county. 20. A band of about 20 is located near Cohagen, and another band of 15 ranges in Townships 13 and 14, both in southern Garfield County. 21. In February, 1924, 38 antelope were reported as ranging on the Timber Creek Divide, a small herd of 7 west of Weldon, another numbering 7 near McDonald Butte, 6 on Jawbone Coulee, and a single buck on the Big Dry, totaling 59 animals in McCone County. 22. Frank Hamlick, a deputy State game warden at Kinsey, Custer County, wrote on January 21, 1924, that 49 antelope were living in his pasture, and that various other bands were located within 15 miles, which in the aggregate amount to about 100 animals. He is doing all he can to protect them, but they are being killed by hunters. 23. Three small bands, aggregating about 19 animals, occur in southeastern Fallon County, where they are reported to be decreasing. 24. A band of 75 is reported to range on Mizpah Creek, in southern Custer and northern Powder River Counties. 25. C. A. Hatterschied wrote in February, 1924, that in the preceding fall he counted a herd of 53 antelope on Timber Creek, in Powder River County. 26. In January, 1924, P. E. Fannigan, of Graham, wrote that several herds of antelope occur in the country lying east of the Big Powder River, in southern Powder River County. He considers that they aggregate more than 300 animals. One neighbor counted 270 antelope in his pasture on one occa- sion last year. They do not appear to be increasing, but some of the farmers complain that they are eating their alfalfa. From the reports, it is probable that there are other antelope in this section of the State, but details are lacking. C. A. Hatterschied reports seeing a herd of 47 on Horse Creek in the fall of 1923. 27. Bands totaling 250 range on the Custer National Forest southeast of Ashland, in Powder River County. B. W. Hogan, of Ashland, wrote that when he went there in 1910 there were only 3 antelope in the entire Custer National Forest. These were well protected by ranchmen and have increased to the present herd. They are often seen in bands of from a dozen to more than 36 BULLETIN 1346, U„ S. DEPARTMENT OP AGRICULTURE 100. Tlie Custer National Forest appears to be ideally adapted to their needs. Powder River County appears to have the largest number of antelope in any area of that size in the State. R. F. Tarbell stated that the largest band on a ranch in the southern part of the county numbered 57. This in- creased somewhat during 1923, but, as a whole, the antelope have barely held their own during the past 10 years. At the present time they are not being molested and undoubtedly are increasing. 28. A band of about 15 is reported in the extreme southeastern part of Big Horn County, on the Tongue River, near Decker. It is being protected and is slowly increasing. 29. A band of about 30, w7hic-h is reported to be increasing in numbers, is ranging south of the Yellowstone on Otter and Beaver Creeks in southern Rosebud County. 30. North of the Yellowstone River, along Stella, Hay, Wolf, and Cotton- wood Creeks, and also between Forsyth and Melstone, are a number of bands of antelope varying in size up to nearly 200 animals, which aggregate about 450. They are said to be holding their own or increasing in numbers. 31. In a district about Melstone, in Musselshell County, three small bands, totaling about 40 animals, are said to be decreasing. 32. A band of 31 is reported on the Gumbo Flats north of Roundup in Mus- selshell County. 33. A band of 172 was counted along Elk River on the Jack Rowley Ranch, about 50 miles southeast of Lewistown, in Fergus County. Mr. Rowley states that for the past 10 years from 100 to 125 antelope have been ranging on the ranch about 50 miles southeast of that town. During the fall of 1923 he counted 172 in one band, which apparently covered the entire number. He states that they ordinarily run in three or four bands, but occasionally unite. They have many young ; but, although efforts have been made by the owners of the ranch to protect them, they continually stray off and are shot by hunters. When fired at, those not hit usually seek safety in the meadows on the ranch where they seem to appreciate the fact that they are protected. Mr. Rowley believes that since so many dry farmers have left that section of the State, conditions are more favorable for the antelope, which are likely to increase in numbers. 34. A band of 7 is located in the foothills of the Big Snowy Mountains south of Moore, in Judith Basin County. 35. About 175 antelope are located mainly in Wheatland County. Of these, two bands of about 20 each range near Rothiemay in western Golden Valley County, about 80 on the Winneoock Sheep Ranch, 5 to 10 miles southwest of Shawmut, and a few small bands numbering about 30 northwest of Twodot, the last-named bands being all in Wheatland County, also 25 near Porcupine Butte, northern Sweetgrass County. It is reported that the antelope in this area have been decreasing rapidly since 1908 through hunting from automo- biles. 36. About 100 antelope ranged near Radersburg, Broadwater County, in 1923. It is stated that some of these can be seen from the main road at almost any time. 37. A band of 12 was reported near Three Forks, in Gallatin County, in 1912, but no information has been received concerning them since. 38. About 22 antelope are reported to live on the old Green Ranch imme- diately west of Madison River near the mouth of Cherry Creek, in Madison County. This band has decreased during the past 15 years. 39. A band of about 10 is located on the Little Timber and Duck Creek Ranges in Sweetgrass County. 40. A band of 10 ranges near Gibson, in the Big Coulee Country, in northern Stillwater County. 41. Several bands, numbering at least 16, range about 20 miles northwest of Billings, in Yellowstone County. 42. A band of 7 is reported on Lone Creek, near Red Lodge, Carbon County. 43. The only antelope reported in Beaverhead County is a band of 8 ranging on Red Rock and Black Tail Creeks. 44. Eight antelope were placed on the National Bison Range in September, 1924, by the Biological Survey, in cooperation with Doctor Brownell, of San Francisco, and Doctor Hornaday, of the Permanent Wild Life Protection Fund. These antelope were from the Washoe Antelope Reservation in north- western Nevada, where they were caught as newly born fawns in the spring STATUS OF THE PRONGKORNED ANTELOPE, 1922-1924 37 of 1924. The antelope placed on this range by the Boone and Crockett Club in 1911 increased to 64, but all are believed to have been killed by predatory animals in 1922. NEBRASKA Of the countless thousands of antelope which once roamed the plains of Nebraska but 10 small bands remain, containing a total of about 187 animals. As in Kansas and some other western States, for many years following the completion of the first transcontinental railroad, passengers through Nebraska had the pleasure of observing many antelope from the train windows. With the increasing occupation of the State by farmers and stockgrowers the prong- horn has been reduced to the present insignificant numbers. Only a single attempt appears to have been made to establish and main- tain under confinement a herd of antelope in this State. This was done in September, 1924, when the Biological Survey, in cooperation with Doctor Brownell and Doctor Hornaday, as detailed elsewhere, placed 10 young ante- lope from northern Nevada on the Niobrara Reservation, a Federal game refuge near Valentine. Conditions there appear to be well suited to antelope, Fig. !). — Distribution of antelope in Nebraska, estimated at 1S7, in 10 areas and it may be possible to build up an interesting herd. In order to insure the perpetuation of these beautiful animals in Nebraska, another herd might well be established elsewhere in the State. The occupation of Nebraska for farming purposes is so complete that there is little hope of a herd being maintained there except under fence. Safe free range for antelope in this State is a thing of the past. In a letter dated March 12, 1924, George Bird Grinnell wrote that three or four years ago a band of about 40 antelope was located somewhere along the North Platte north of Bridgeport, Morrill County. According to the latest accounts he had they were rapidly decreasing. Supervisor Jay Higgins, of the Nebraska National Forest, supplied in- formation in the spring of 1922 that there were bands of antelope "in Scotts Bluff, Banner, and Kimball Counties, and added : “ We secured three con- victions for killing antelojie in Scotts Bluff and Kimball Counties.” - The antelope bands existing in Nebraska in 1922 and 1926 were distributed as follows (fig. 9) : 1. A band of about 12 reported in 1922 as near the 33 Ranch, in Sioux County, near the Wyoming line. 38 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE 2. A band of about 25 ranged near Agate, in Sioux County, in 1922. These have existed in about the same numbers for the past 10 years. 3. A band of about 5 was reported in 1922 about 10 miles west of Bushnell, in Kimball County. 4. A second band of 5 was reported in 1922 near the State line, south of Kimball, in Kimball County. 5. A band of 14 was reported in 1922 between Dix and Potter, on the border between Kimball and Cheyenne Counties. Their numbers for some years have remained about stationary. 6. A band of about 40 was reported in 1922 as ranging about 18 miles south of Sidney, in Cheyenne County. For some time this band has about held its own. 7. A band of 43 was reported in the spring of 1922 as grazing in fields near Sunol, in eastern Cheyenne County. The farmers complained of this invasion of their wheat fields by antelope and requested information of the Forest Service as to what might be done to control them. 8. A band of 25 was reported in 1922 about 12 miles south of Lisco, Garden County. This herd had about held its own for some time. 9. A band of 8 was reported in March, 1924, about 15 miles north of Sargent, in Loup County. 10. Ten young antelope, 6 females and 4 males, were placed on the Niobrara Game Reservation in September, 1924, by Doctor Brownell, of San Francisco, and Doctor Hornaday, of the Permanent Wild Life Protection Fund, in coopera- tion with the Biological Survey. These antelope were part of the fawns cap- tured in northwestern Nevada in the spring of 1924, as already detailed. NEVADA Formerly antelope were plentiful over all the Great Plains and deserts in this State. They are now limited to about 11 comparatively small areas. In the northern and northwestern parts of the State occur great areas sparsely occupied by man, where conditions are still favorable to these animals. As a result, some of the largest herds to be found in the West still survive there, aggregating more than 4,200. Legal protection for antelope in Nevada has undergone some changes in the past few years which it will be of interest to record. The close season for antelope was lifted in Nevada by an act approved March 23, 1921 (amending chapter 234 of the act of March 27, 1917, as amended by act approved March 4, 1921), providing as follows: “ Sec. 42. It shall be unlawful to kill, catch, trap, wound, or pursue with the intent to catch, capture, injure, or destroy any deer or antelope at any time during the year other than during such 30-day period to be known as the open season between September 15 and December 15 of each year as may herein- after be designated for the respective counties by the boards of county commis- sioners thereof under the provisions of section 50 of this act ; provided that there shall never be any open season on deer without horns and that during such open season of each year it shall be unlawful to kill, catch, trap, wound, or pursue with the intent to catch, trap, injure, or destroy more than one deer with horns and one antelope with horns ; and provided further that in all counties in which no designation to the contrary shall have been made by the county commissioners prior to the 1st day of August of any year, the open season for deer with horns or antelope shall be from October 14 to November 12, both dates inclusive.” An act approved March 21, 1923, restored antelope to the protected list until 1930, in the following terms : “ Sec. 9. It shall be unlawful at all times to take any mountain sheep, goats, elk, or antelope until January 1, 1930.” After the opening of the season on antelope in Nevada in 1921 there was a feeling among some of the county commissioners in the northern part of the State that an open season was called for. E. R. Sans, predatory-animal inspector of the Biological Survey, working with the Washoe County Game STATUS OF THE PRONGHORN ED ANTELOPE, 1922-1924 39 Protective Association, pointed out to the boards of supervisors in Humboldt and Elko Counties the undesirability of an open season on antelope, and as a result the season was closed by them for two years. In Washoe County the board of supervisors felt the technical requirements of the law called for an open season, but were so well convinced of the undesirability of permitting antelope to be killed in any numbers that they limited the open season to one- half hour, extending from 7.30 to S a. m., November 10. It is obvious that under such conditions no antelope were legally killed in Nevada that season. The information concerning the distribution and numbers of antelope in Nevada lias been furnished mainly by Mr. Sans and by Supervisor Alexan- der McQueen, of the Hum- boldt National Forest. Mr. Sans has been extraordi- narily successful in pro- moting antelope protec- tion in Nevada. His friendly cooperation with the State authorities and the active part be took at the request of the governor in helping to locate and outline the antelope ref- uges in the northern and northwestern parts of the State have been a major service to the conservation of these animals. The distribution of an- telope in Nevada is ap- proximately as follows (fig. 10) : 1. This area is the Washoe County State Game Refuge (PI. V, fig. 2). The number of ante- lope within its limits is estimated by resident stockmen at from 2,000 to 2,500. E, R. Sans wrote : “Predatory Animal Hunter R. W. Young, sta- tioned at the Tbomas Du- furrena ranch in the Fig. 10. — Distribution of antelope in Nevada, estimated Thousand Creek district at 4,253, in 11 areas. Game refugees indicated by of Humboldt County, re- broken lines ported on January 17, 1924, that he saw not less than 1,000 antelope in the course of a day’s travel in that neighborhood, and from reliable reports we gather that this is the wintering ground for these animal's. “ T. B. Harriman, one of our predatory-animal trappers working in northern Washoe County, reported December 10, 1923, a band of from 600 to 700 ante- lope migrating to their winter range in the High Rock Canyon. This is the largest band we have had any notice of wintering in High Rock Canyon. There have always been a few coming into this sheltered district to winter, but this winter an unusually large number is there.” Further interesting information concerning the antelope on this game refuge is contained in statements received from Mr. Sans, which are set forth in the account of bis capture of the young antelope for restocking purposes during the present season, 40 BULLETIN 1346, U. S. DEPARTMENT' OP AGRICULTURE 2. This area covers the Smoke Creek antelope refuge of Washoe County, south of the main Washoe County refuge. Mr. Sans wrote : “ From reliable information I learn there are about 1,000 antelope ranging from Willow7 Creek, northeast of Susanville, in California, to Smoke Creek, in Nevada. The larger part of these appear to range in Secret Valley and the tablelands in Nevada to the railroad to Amadee and Ravendale, Calif., on the north.” These herds include those recorded for Lassen County, Calif., and those which remain permanently in Nevada. On account of the uncertainty as to the exact number in Nevada, they have been placed at 200 animals, although at present they must exceed 1,000, owfing to the California herds having temporarily entered this area, as set forth in the account of the California antelope. 3. About 40 antelope are reported to range on the Santa Rosa State Recre- ation Ground and Game Refuge in eastern Humboldt County. 4. Various bands, aggregating about 1,000 antelope, are reported to occupy this area, which includes the Humboldt State Recreation Ground and Game Refuge, in Elko and Humboldt Counties. This area is the southern extension of the Owyhee Desert from across the boundary in Idaho. Some of the antelope range back and forth across the Idaho line. These bands are said to be holding their own, if not increasing. 5. This area contains bands numbering, respectively, 29, 43, 71, and 70, by actual counts, ranging on Nine Mile Flat, 16 miles east of Contact and be- tween the Bad Lands and Loomis Creek, in Humboldt County. These 213 are said to have increased from 20 during the last seven years. 6. A band of about 10 ranges near Cobre, in Elko County. 7. This area covers the White Pine State Recreation Ground and Game Refuge (No. 12), in White Pine and Elko Counties. A band of 40 antelope is reported to be ranging there. 8. A band of about 75 ranges in Duck Valley, from Geyser to Pioche, in Lincoln County. 9. This area includes the Grant State Recreation Ground and Game Re- fuge (No. 4) in Nye County.. Several small bands of antelope, estimated to aggregate from 35 to 65 animals, are said to range within this area in Railroad Valley. 10. A band estimated at 100 u7as seen during the spring of 1923 near White Blotch, Lincoln County, and in the adjacent parts of Nye County. 11. A band of 25 is reported to range in Wild Horse Valley, southern Nye County. NEW MEXICO Antelope in New Mexico are decreasing, but up to the fall of 1923 they were still found in 31 areas, with an estimated total of 1,682 survivors from the vast herds which once occupied this region. Details concerning their numbers and distribution in this State set forth below are largely the result of careful in- vestigations made to March 1, 1924, by L. C. Petree, chief deputy in the State department of game and fish. In addition information has been supplied by employees of the Forest Service and of the Biological Survey and by individuals in the State. District Forester Frank C. W. Pooler, of Albuquerque, submits some interesting ideas on antelope conservation, as follows : “ I imagine everybody agrees that the nucleus of any scheme should include several Federal game preserves covering herds like the one proposed in adja- cent parts of Oregon and Nevada. Such preserves, however, can not go fur- ther than to serve as a kind of rock-bottom insurance against total disappear- ance. The big problem is to secure an effective care of the scattered herds running on all kinds of land under all kinds of jurisdictions. “ Could not the Biological Survey be designated by the proposed convention as the central agency to perform the following steps with respect to each herd for which there appears to be a reasonable chance of perpetuation : “(1) Determine the number, range, and condition of the herd. “(2) Assign custodianship of the herd to some one party. This might be the Forest Service, the State game department, some stockman, or possibly some game protective association. STATUS OP THE PBONGHOENED ANTELOPE, 1922-1024 41 “(3) Confex* with such agency as to the step necessary to insure perpetu- ation ; such steps might include the establishment of State game refuges, the offering of rewards against killing, pledges on the part of the stockmen to exer- cise rigid jurisdiction over their employees, pledges of winter feeding, or arrangements for predatory-animal control. “(4) Ask for the necessary cooperation from the necessary parties to as- sure the execution of the measures decided on under the preceding para- graph (3). “(5) Require from the custodian at least an annual report on the condi- tion and needs of the herd. “ Let me offer one example of how the foregoing scheme would work. On Diamond Creek in the Gila Forest is a herd of about 25 animals, which have lodged there in timbered country since a blizzard drove them out of the San Augustine Plains about five years ago. This herd has plenty of feed and water, very fair protection against predatory animals, and no great likelihood of illegal killing except during the deer season, when they are occasionally mistaken for deer by hunters. A special warning to all hunters in that locality, supplemented by a little extra patrol during the 10 days of the hunting season, would, I think, cause them to start increasing. The Forest Service would, I think, assume the special custo- dianship of this herd and would have the cooperation of the stockmen. If au- * thorized by the forester it might be that we could even require certain special precautions by the two or three stockmen affected against any of their em- ployees damaging this herd. Should predatory animals1 get worse, we would doubtless have your coopera- tion whenever we reported the situation. The main thing would be that some- body would assume responsibility for doing all reasonable and practicable things for the herd.” Mr. Pooler’s contribution contains some very practical suggestions, some of which might be utilized as the program for antelope conservation develops. Antelope for many years have been protected on the well-known Bell ranch, where there has been a standing offer of $50 reward for the arrest of anyone found hunting on these lands ; but under this protection they have not in- creased so rapidly as might be expected, probably owing to the depredations of predatory animals and eagles. The bands of antelope in New Mexico are located as follows (fig. 11 ) : 1. About 50 antelope range in southeastern Colfax County. 2. A band of 14 was reported in February, 1024, in the Eklund pasture, 15 miles northwest of Clayton, Union County. Fig. 11. — Distribution of antelope in New Mexico, esti- mated at 1,682, in 31 areas 42 BULLETIN 1346, IT. S. DEPARTMENT OE AGRICULTURE 3. About 300 antelope are estimated to occupy the area of the original Bell ranch, eastern San Miguel County, now partly owned by the Tom B. Owens Co. and by Dan C. Trigg, jr. This is the largest number of antelope in any re- stricted district in the State. The first-hand information here presented con- cerning these animals shows the need of a careful survey of the situation in this district as a basis for further conservation measures. The friendly atti- tude of protection toward the antelope by the owners of the ranges indicates possibilities of building up here considerably larger herds than now exist. Owing to the large size of this area there has been some difficulty in getting definite information concerning the present situation. C. M. O’Donel, manager of the Bell ranch, under date of July 20, 1923, supplied the following : “ The sum of the reports from employees in various parts of the range gives the number of antelope within the present boundaries of the ranch as 217. Naturally this can not be an accurate count, though the habit of antelope to ‘ locate ’ in bunches makes it more accurate than would probably be the case with other varieties of game. * * * “ I believe that antelope are increasing on this range only very slowly, if at all. My opinion is that their natural enemies, of which perhaps the eagle is the worst, keep down the increase by destroying the young. I believe we had as many, if not more, antelope here when I first came to the ranch 25 years ago.” Inquiry was instituted among the purchasers of parts of the original Bell ranch, with the following results: The Tom B. Owens Co. wrote : “ We hardly know how to arrive at an estimate of the number of antelope on our property, but think around 100 to 150 old ones, with possibly a fawn crop of 50 this year. “ The antelope on our place as well as on the Bell ranch are found on the level open valleys and rarely go into the mountains for any reason. We have, it seems, a surplus of bucks and often see them off by themselves, they having been whipped out of the herds by the younger and stouter bucks. “We do not know of any other section of the country where the antelope are as often seen as on these ranches, and we never take any kind of drive or ride over our pastures without seeing several bunches of various numbers, from 2 to 20.” Dan C. Trigg, jr., who now owns a part of the original Bell ranch, wrote : “ A few antelope stay in my pasture all the time. I have seen as many as 26 in a bunch. There have always been two separate bands in different portions of my ranch. They are more or less migratory and have a habit of crossing into the Bell ranch, which joins my holdings for several miles.” 4. Bands totaling about 60 antelope are reported from the ranches of Senator A. A. Jones and of John Hicks, in San Miguel and Guadalupe Counties. 5. A band of 50 is reported in western Guadalupe County. 6. Three antelope are on the Ed. Morrow ranch, in southern Guadalupe County. 7. A band of 10 is reported in eastern Guadalupe County. 8. Four antelope are reported on the Buckeye ranch, near Taft, in northern De Baca County. 9. A band of 12 is on the Charles Orr ranch, near the cornering parts of Roosevelt, Curry, and Quay Counties. 10. A band of 8 occurs on the C. S. Hart ranch, near the borders of Roose- velt and Curry Counties. 11. A band of 12 is near the bordering corners of Lincoln, De Baca, and Chaves Counties. 12. A band of antelope is reported as living in southeastern Socorro County, but the number is not given. 13. Herds aggregating 200 or more are reported as ranging on the San Augustine plains, in Catron and Socorro Counties. This number was verified by actual count reported by the Magdalena Game Protective Association, but those familiar with the situation believe that there are many more than that number in this district. G. W. Evans, of Beaverhead, states that 200 antelope, by actual count, live on his 50,000-acre ranch in the southwestern portion of San Augustine plains, in Catron County, within the general area reported by the Magdalena Game Protective Association. It is obvious that there are in this Great Plains region many more antelope than those here listed, possibly 500 in all. Formerly the San Augustine plains were a favorite resort for thousands of antelope. 14. A band of about 60 ranges in western Sierra County, STATUS OP THE PRONGHORNED ANTELOPE, 1922-1924 43 15. Clyde L. Grow, reservoir superintendent of the Reclamation Service, wrote from Engle, N. Mex., on September 12, 1924, reporting between 50 and 75 antelope on the east side of the Elephant Butte Reservoir, in the Cristobal Mountains, in eastern Sierra County. He added that they were fed during the deep snow all the preceding winter by the Victoria Land & Cattle Co. and are in good condition. They range to the vicinity of Engle, where they are sometimes seen by passengers on the Santa Fe Railroad trains. In 1883 the writer had the opportunity to observe personally a very considerable number of antelope ranging the plains about Engle, particularly to the east and north. The proprietor of the single hotel and general store there at that time kept a pack of greyhounds which, he fed on antelope meat. His sole amusement in this isolated place was to drive out with a buckboard on the open plains, accompanied by his greyhounds, until he found a band of antelope, when the greyhounds were sent in pursuit while he followed until the dogs had pulled down and killed one or more of the animals, which he carried back for dog food. 16. A band of 56 antelope was reported to be ranging between White Sands and the Organ and San Andres Mountains, in Socorro and Dona Ana Counties. 17. In southwestern Otero County 90 antelope are reported by Oliver Lee, manager of the Sacramento Land & Cattle Co., and others. 18. A band of 25 ranges about the headwaters of the Felix River, in southwestern Chaves County. 19. The “ L. E.” pastures in eastern Chaves County are occupied by a band of 15. 20. H. E. Crosby, of Ivenna, reports 40 antelope living in the pastures of the Crosby ranch in eastern Chaves County. 21. A band of 30 is reported ranging on the Littlefield ranch, on the Staked Plains, in central Roosevelt County. 22. In northeastern Lea and southeastern Roosevelt Counties 45 antelope are reported ranging on Bakers Flats and across into adjacent parts of Texas. A small band, the number not specified but reported to have been living in northern Lea County for several years, has raised no fawns, owing to the depredations of predatory animals. In the spring of 1923, following a co- operative campaign against these destructive pests by the Biological Survey and the State, a number of fawns survived and this herd may now increase. 23. A band of 35 antelope is reported in western Lea County. 24. There is a band of 10 near Cow Springs, in southern grant County, and one of 22 on the Antelope Plains of western Luna County. 25. A band of 20 is reported as ranging in the San Luis Valley, in Hidalgo County. 26. A band of 7 is in southwestern Luna County. 27. A band of 7 is reported also in the Juniper pastures, Animas Valley, in Hidalgo County. 28. In 1922 a band of 12 was reported in Playas Valley, in Hidalgo County. 29. About 40 antelope range in extreme southeastern Hidalgo County, some of which cross into the adjacent part of northwestern Chihuahua. 30. Twenty antelope range from southern Otero County south into Texas. 31. About 50 antelope occur in San Simon Range, in southern Lea County. NORTH DAKOTA Antelope have almost disappeared from North Dakota. The remaining herds now number only five and aggregate 225 animals. Their future appears to be extremely doubtful unless a game preserve can be established wherein they may be safeguarded. The information concerning antelope in North Dakota has been obtained by H. L. Rice, of the North Dakota State Game Com- mission, and R. Scott Zimmerman, in charge of rodent-control work in the State for the Biological Survey. The distribution of the herds is approximately as follows (fig. 12) : 1. In September, 1924, 00 antelope were reported as ranging from north- western Dunn County into the adjacent part of McKenzie County. 2. A band of 9 was reported in September, 1924, in southwestern McKenzie County. 44 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE 3. About 75 are reported in adjacent parts of central Golden Valley and Billings Counties. Tbis is the largest band reported in tbe State. William McCarthy, who owns 11,000 acres of rough, rolling land in the heart of the Bad Lands along the Missouri River, which affords a natural range for game, writes that when he came into possession of the range in 1910 there were about 15 antelope there. Much hunted, they sought and were given every protection in his pastures, where they found running springs and flowing wells with an abundance of grass, and as a result have become very tame. 4. Bands numbering 55 were reported in September, 1924, in the Bad Lands of the Little Missouri River in Slope County. 5. In September, 1924, a band of 26 was reported from southwestern Bowman County. OKLAHOMA Of the vast number of antelope once roaming the prairies of Oklahoma only a single native band, containing 5 or 6 animals, was reported as surviving in 1923, and the small band on the Wichita National Game Preserve, in Comanche County. (See PI. I.) NORTH DAKOTA Fig. 12. — Distribution of antelope in North Dakota, estimated at 225, in 5 areas In December, 1910, and January, 1911, the Boone and Crockett Club trans- ported 9 antelope from the Yellowstone National Park herd to the Wichita National Game Preserve. This experiment had an unfortunate ending, since all the animals died during the next few years. Another attempt was made in the fall of 1921 by the American Bison Society to established a herd on this preserve by placing there 10 animals which had been purchased at Brooks, Alberta. Six of these died shortly afterwards, and in the fall of 1922 the Bison Society placed 6 more there from the same source. Of these 5 died shortly afterwards, leaving during the winter of 1922-23, 5 survivors from the original transplantings of 16. In the spring of 1923 the 3 females each gave birth to a pair of young, which were^ safely reared. This was duplicated in the spring of 1924, bringing the number in the herd to 17. The handicap which at first existed appears to have been overcome, and the outlook is favor- able for the establishment there of a good herd. 45 STATUS OF THE PR-ONGHORNED ANTELOPE, 1922-1924 The following interesting quotation from a letter from District Forester Reed, United States Forest Service, dated June 20, 1923, gives an idea of the vicissitudes undergone by the antelope during the last two introductions : “ We have just received word from Mr. Rush that there is but one survivor of the six shipped to the Wichita last fall. This survivor is a buck. Two of the antelope died from the effects of ticks and two have disappeared. Mr. Rush surmises that the coyotes got in and killed them while they were in the little bull pasture. Later they were moved into the buffalo yard, and the only female left ran headlong into the gate and broke her neck. Of the antelope shipped two years ago, one 2-year old buck and three 2-year old does remain. This reduces the herd to 5 adult antelope. “ Mr. Rush reports that the 3 does now have 2 fawns each. This brings the herd up to 11 head, and Mr. Rush says that he had excellent luck with them. It is to be hoped that the fawns born in captivity on the Wichita will survive the vicissitudes which decimated the original shipment made by the American Bison Society. Since we have 11 antelope on the Wichita, it does not seem necessary to seek further assistance from the American Bison Society at this time. We will, however, take the best care of the remaining antelope. We are satisfied that Mr. Rush has done his best, and it seems that we have a fighting chance to secure a herd of antelope on the Wichita.” The location of the two bands of antelope now in Oklahoma may be stated as follows (fig. 13) : 1. A single band of 5 or 6 animals is living in the Ford pastures in northeast- ern Cimarron County, where it is protected. A band of about 8 animals, reported to range in Morton County, southwestern Kansas, is said to spend part of its time across the line in Oklahoma, which would take it into Cimarron County. The relations between these two herds have not been ascertained. For convenience the 8 animals are credited to* Kansas and make up the only known surviving antelope in that State. Apparently the only survivors of these animals in both Oklahoma and Kansas are in the adjoining counties of the extreme western parts of these States. 2. During tbe summer of 1924 a herd of 17 antelope was on the Wichita National Game Preserve. This is an increase of 12 animals from the 5 survivors of 25 animals imported in previous years by the Boone and Crockett Club and the American Bison Society. 46 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE OREGON Southeastern Oregon forms part of a rough, rocky desert covering also northern Nevada and southwestern Idaho, on which natural conditions have been exceedingly favorable for antelope. This region constitutes one of the few areas in the United States where large herds of these animals numbering hundreds still continue to congregate during the winter season. Southeastern Oregon covers so large a territory and the herds in it are so widely scattered that it has not been practicable definitely to locate them and ascertain their numbers. For this reason an area has been marked on the accompanying map (fig. 14) covering the main antelope territory, within which it is esti- mated that the different herds contain an aggregate of not less than 2,000 animals. Most of the definite information concerning antelope in Oregon has been supplied by Stanley G. Jewett, of the Biological Survey, and W. L. Finley, of the National Association of Audubon Societies. There is no question that antelope have increased in Oregon during the past few years, and although year by year a considerable number have been killed by poachers, yet this has not been sufficient to overcome the increase. It has been reported that the climatic conditions were especially favorable for them during the winter and spring of 1923-24, and that an unusually large number of young were bom. Water and range conditions were worse in this district during the summer of 1924 than for years. Cattle owners moved all their stock from this range about the middle of August and as a result con- ditions were made more favorable for the antelope. Old Fort Warner and the neighboring Desert Lake appear to be centers of abundance for antelope. Stanley G. Jewett, leader of the predatory animal control work of the Biological Survey in Oregon, writes that while he was there during August, 1924, antelope were in sight practically all the time, and he is confident that on August 16 and 17 he saw not less than 500 within a radius of 15 miles. In a letter dated September 2, 1924, Mr. Jewett stated : “ I am sorry to report that a number of fawns have been found dead. Jacobs reports about 20 dead within a radius of 15 miles from old Fort Warner. An old doe was sick near camp while I was there. She acted much like an alkalied cow. This condition has probably been brought about by the extreme drought and the fact that the does have not had enough nourishment properly to feed their young. Range conditions are so bad that the big cattle companies have taken all their cattle from that range.” Such adverse conditions must prevail not only over eastern Oregon but into the adjacent parts of Nevada and Idaho. What the outcome will be as to the antelope in this great area is a serious question, since it is one of the greatest centers of surviving antelope in the entire West. For a number of ' years various persons interested in the conservation of antelope have been advocating the establishment of a Federal antelope refuge in southeastern Oregon. Details concerning this project are set forth else- where in this bulletin. The present distribution of antelope in Oregon is as follows (fig. 14) : 1. During July, 1924, a single antelope was seen at different times near Antone, in Wheeler Caunty, which is considerably outside the general distri- bution area of antelope at the present time and may indicate a gradual exten- sion of range into formerly occupied territory. 2. In July, 1924, an isolated herd of about IS was observed on Twelve Mile Creek in the southeastern part of Crook County and the northwestern part of Harney County. 3. A herd of about 20 was ranging in July, 1924, on the northeastern side of Harney Valley from Saddle Butte north to old Camp Harney .a northern Harney County. 47 STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 4. This large generalized area contains nearly all the surviving antelope in Oregon. The number has been roughly estimated here at 2,000, although it may be considerably in excess of this. They are distributed in many herds, the largest of which is believed to number nearly 1,000 and is located in south- ern Harney County and southeastern Lake County, from which it may range across into Nevada. Two other herds, estimated to contain about 500 each, range, one in southern Malheur County across the boundary into Idaho and into Nevada, and the other in northern Lake and southern Deschutes Counties. The many dry-farming homesteaders have left the high desert country of southeastern Oregon during the past few years, and the antelope have been gradually going back into their former range. During the summer of 1924, seven grown antelope and one fawn were seen a number of times in the jack- pine timber about Button Spring and Sand Spring, where they watered, in the extreme northeastern corner of the Deschutes National Forest. Another recent extension of range is in the vicinity of Fife in southern Crook County and along the upper stretches of Silver Creek in northwestern Harney County. Fig. 14. — Distribution of antelope in Oregon, estimated at 2,039', in 4 areas There small bands of from 10 to 20 have been seen at various times about Benjamin Lake and on Wagontire Mountain. The antelope occurring in the northwestern part of their range in this region are scattered in small bands, owing to the extreme scarcity of water. They practically all water at the springs on Grays Butte, Christmas Lake, Button Spring, Sand Spring, and the southern slope of Hampton Butte, and ordinarily at Glass Butte, but the latter spring has been absolutely dry the present sea- son. Antelope in the extreme southeast bordering the Idaho line are in a better watered region and are much scattered along the tributaries of the Owyhee and about many springs in that area. SOUTH DAKOTA Senator Peter Norbeek and State Game Warden II. S. Hedrick are taking a very active interest in the conservation and building up of the herds of antelope in South Dakota. Concerning the practical side of this question, Senator Norbeck’s remarks at the antelope conference in Washington on December 14, 1923, are much to the point. He stated: 48 BULLETIN 1346, U. S, DEPARTMENT OE AGRICULTURE “ I think the situation in our State is very largely the same as in other Western States. The antelope is exterminated everywhere except in about one-quarter of the area. Together with the State game warden I spent a little time this summer going over 5 or 6 counties and we were surprised that there were a number of small bands of antelope surviving. They remain in certain areas which are probably more favorable for them. The bands were generally from 4 or 5 up to 20 or 30. One band of 85 was seen — a really fine-looking, healthy lot of animals. They had been ranging in the same neighborhood for about the last 20 years. * * * “ South Dakota is all settled. There is very little Government land left. The land we need for the antelope refuge is nearly all patented, though not all occupied. The State has a fenced game preserve of 40,000 acres, but this is built in the foothills of the mountains and is not a suitable range for antelope. While I was there recently the State game and fish commission passed resolutions taking the first steps toward the establishment of an antelope preserve in the antelope country, with the plan of fencing in 5 or 6 sections of land to include some of the larger bands that we saw during my recent trip. This should take in from 50 to 100 antelope as a start. I am sure that something substantial will come from this.” Senator Nor beck informs the writer that this game refuge will be primarily for antelope, but that with the addition from the Federal forested lands it is desired to establish here herds of elk, buffalo, and possibly some other game animals. The headquarters of this fine game refuge is to be at Reva Gap, located on the main line of an important highway. This locality is not only one of natural beauty but one of historical interest, having been the scene of the battle of Slim Buttes with the Indians 50 years ago, fought under Gen. Anson Mills, then a captain. Parts of this game preserve are hilly, with thin forests ; the rest of it is open prairie. The largest number of surviving antelope are located in the northwestern corner of the State, where, in Harding County, a new State antelope refuge has been established in accordance with legal authorization granted at the time Senator Norbeck was governor. In regard to the plans for this refuge Senator Norbeck wrote under date of July 24, 1924: “An antelope preserve has been established in the northwest corner county of the State by action of the State game and fish commission, and additional Federal lands have been set aside for the purpose by recent act of Congress. “ The area includes considerable State land, but some private ranches will have to be purchased. The plan is to have an inclosed preserve of about 15,000 acres. The first fence, which is now under construction, incloses an area 3 miles square. It is believed that from 100 to 150 antelope can be gathered into this inclosure, as that number of animals range over this area and in the immediate neighborhood. “ The State has set aside $20,000 for this work. Additional funds will be required, but same will be provided in the next few years. It will probably take from 3 to 5 years to work out the complete plan but I believe that we have made a very good start." On August 8 Senator Norbeck wrote that after further consideration on the ground it has been agreed immediately to enlarge the fenced area on the new antelope refuge to include 15 or 16 sections of land. Under date of September 9, 1924, he added : “ We are going ahead in splendid shape with our antelope preserve. The material has already been purchased for the inclosure of some 15 or 16 sections STATUS OF THE PRONGHORN ED ANTELOPE, 1922-1924 49 of land. Tlie fenced area will be approximately 4 miles square and will cover some of the present antelope range. “ Whether it will be 15 or 16 sections depends on the purchase of a ranch, for which negotiations are now under way. Most of the land inside the in- closure is owned by the State of South Dakota. An 800-acre ranch, with im- provements, located in the center of the area, has already been purchased by the State. “ The inclosure will cover approximately half of the proposed game preserve, it being the intention of the commission to enlarge it in a year or two by add- ing an area 4 miles square, which will include a few sections of forest-reserve land in the vicinity of Slim Buttes. “ The preserve is located in the eastern part of Harding County and is 84 miles from the closest railroad point by present highways. This, of course, makes the undertaking rather expensive; but it is a splendid location, even though somewhat isolated.” In connection with the establishment of the State antelope refuge in South Dakota, mentioned above, the following letter, dated December 4, 1928, from State Game Warden Hedrick, is worth quoting: “ Senator Norbeck and myself have been making a personal investigation along this line, having recently put in several days in Harding County, in the northwest corner of South Dakota, investigating conditions and looking for a location for the establishment of an antelope preserve, which was author- ized by the South Dakota Game and Fish Commission during the time that Senator Norbeck was governor of the State. “ When the Senator arrives at Washington he will doubtless see you per- sonally and paint a word picture to you of this beautiful prairie animal, as he certainly got very enthusiastic when we came upon a band of 85 head on a fine Sunday afternoon and were within 200 feet of a considerable number of these animals at times. Within 3 miles of this place on the same after- noon we came upon another band of 17 and drove up within 8 or 10 rods of them. Thei'e was also a band to the west of us that we did not get close to ; we do not know how many there were in this band. Upon talking to the neighbors and ranchers in that section, where the antelope seem to have many friends, my estimate would be that there are at least 225 antelope within a range of 4 to 6 townships. There are also many other bands in Harding County, as well as in Perkins, Butte, and Meade Counties. We also have a band of from 50 to 75 head within 50 miles of Pierre, lying to the northwest of us, in the Cheyenne River country. The Senator and I investi- gated this situation the latter part of July, this year.” Three unsuccessful efforts have been made to stock the large State game park of South Dakota, but in each case the animals died from disease or other causes. The new antelope refuge is in much more suitable country, and there the animals should do well. On May 20, 1923, Louis Knowles, predatory-animal inspector of the Biological Survey, who furnished most of the information as to the specific distribution of antelope in South Dakota, wrote that he believed antelope have decreased 50 per cent during the year. This has come about through depredations of predatory animals, diseases, illegal shooting, and a shortage of males. Coyotes are reported to kill many antelope. One of the official hunters has been work- ing in the principal antelope ranges, where he has killed many of these preda- tory animals, thereby relieving the herds from one of their chief dangers. On June 23, 1923, Mr. Knowles wrote that stockmen and others throughout the country where the surviving antelope occur report a marked decrease in 50 BULLETIN 1346, U. S. DEPARTMENT OP AGRICULTURE their numbers during the preceding 12 months. The only exception to this is in Harding County, where an increase was reported for the past two years, this possibly being due to animals having come in from other sections. A number of small bands of antelope have been exterminated in the State within the past few years. Stanley Phillips, present owner of the Phillips buffalo herd, informed Mr. Knowles that antelope in northern Stanley County were rapidly decreasing. He reported the existence of a good-sized band there two years ago, which has since been hunted with dogs and has been rapidly depleted. It is reported that officers who were searching the premises of an alleged “ moonshiner ” in Harding County found 11 antelope skins. It is encouraging to learn that the people in the town of Buffalo are organizing a rod-and-gun club largely for the purpose of giving protection to the remaining antelope in the State. Fig. 15. — Distribution, of antelope in South Dakota, estimated at 680, in 11 areas Mr. Knowles wrote that there has been a disproportionate decrease in the number of buck antelope, and one of the small surviving bands is composed entirely of females. Owing to the scarcity of males throughout the antelope country many of the females do not breed. In one band of 40 only 3 bucks were found. On June 29, 1923, J. D. Carr, writing from Lindsay, stated that 75 antelope range within a radius of about S miles in the Cheyenne Breaks, where they are not being molested. On the same date, from the same locality, F. L. Norman wrote that about 125 antelope are running near Lindsay, where they are so tame that they often come within 100 yards of his home. The crop of young for the season appears to have been large. Mr. Norman states that he and his son try to protect the antelope in every possible way and that they will be pleased to have any measures taken to insure the safety of the herd. The remaining antelope herds of South Dakota appear to be distributed in the following 11 areas (fig. 15) : STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 51 1. A band of 12 was reported in June, 1923, in western Harding County, probably ranging across the boundary into Montana. 2. In November, 1923, 18 antelope were reported near Bison, in Perkins County. 3. The largest herds of antelope in the State were reported in 1923 by Senator Norbeck and by the Biological Survey representative, Louis Knowles, as existing in adjacent parts of Harding, Perkins, Butte, and Meade Counties. About 150 animals were reported from around Bam Butte. During this same year O. W. Litzke reported having seen about 300 in a 20-mile ride in the Slim Buttes region. In November, 1923, Senator Norbeck wrote there was a herd of 32 in a pasture 3 miles northwest of Camp Crook, and another small herd of S in a pasture near Reva Gap, east from Buffalo. Several scattered small bands occur about the north end of Slim Buttes and several large bands about 15 miles southeast of Buffalo, in the neighborhood of Bam Butte, being some 6 miles west of Slim Buttes. Senator Norbeck added : “ I saw two small bunches on a quarter section of land, one of about 40 animals and the other about 45. Several other bands containing from a dozen to two dozen animals each were seen in a couple of miles in different directions from the larger bands. In other words, there must be from 125 to 150 antelope within 5 miles of Bam Butte, which is a small but well-known landmark in the neigh- borhood.” A reasonable estimate of the total number of antelope in this area is 350. 4. Thirty-seven antelope are reported to be living on the Belle Fourche Bird Refuge, in southwestern Butte County, in the Paul Bernard pasture, where they are being protected by the owner. 5. About 125 antelope are reported in northern Stanley and eastern Haakon Counties. Concerning these, Senator Norbeck wrote, under date of July 14, 1924, that he is well pleased with a trip made recently into the Cheyenne River country, and that “ careful inquiry among the ranchers who1 have had a friendly attitude toward the antelope convinced me that there were probably 100 animals ranging over an area 4 miles wide by 10 to 12 long in the breaks on the south side of the Cheyenne River. The distribution seems to be about equal between Stanley and Haakon Counties. It was a surprise to me to find the antelope here, as it did not appear to be choice antelope range, but they have existed here for about 30 years and apparently have held their own.” A band of 30 is probably permanently located on the Carr ranch, in north- eastern Haakon County, where there is a pasture about 2 miles wide by 4 miles long, in which the antelope range most of the time, although a year ago last winter they spent several weeks, if not months, on the river fiat in an alfalfa field on the ranch. The owner states they did no damage to the alfalfa. Usually they range in the hills and are often seen on high points. Louis Knowles wrote that some appear to be very wild, in part due to the hunting of predatory animals with dogs in this district, during which the dogs fre- quently pursue the antelope. In addition, there has been a certain amount of hunting with guns. A local hunter agreed that the antelope have not increased here for several years, but during the past three years have about' held their own. G. About 100 antelope were reported in adjacent parts of Stanley and Hughes Counties, where they were decreasing rapidly through being hunted with dogs. 7. Twelve antelope were seen near Scenic, in southeastern Pennington County, in March, 1922, by H. R. Wells. 8. Only 3 antelope, all females, were reported to survive on the Pine Ridge Indian Reservation, in Washington and Shannon Counties. 9. A band of about 30 was reported in 1923 from western Fall River County. 10. In August, 1924, a bunch of G antelope was seen 11 miles west of Ardmore, Fall River County. Antelope have supposedly been extinct in this locality for several years, and it is thought this bunch must have drifted in from the west. 11. In October, 1914, 13 young antelope captured near Brooks, Alberta, were placed on Wind Cave Reserve, the gift of the Boone and Crockett Club, of New York City. (See PI. VI.) Another shipment of 9 animals from the same source was received in October, 1916. The antelope increased very well, but losses were great, caused partly by sickness and partly by attacks of coyotes. Coyotes have been a source of much trouble and in 3918 killed 13 antelope here. Trappers have been sent to the preserve at various times to assist in exterminating these and other laudatory animals and have killed 52 BULLETIN" 1346, U. S. DEPARTMENT OF AGRICULTURE a large number of them. The herd is still more or less in danger, however, from attacks of their predatory enemies. It was reduced to 8 animals in December, 1915 ; increased to 23 in 1917, and to 34 in 1921 ; but was again reduced during 1922 and 1923 to 17, and during 1924 to 6 animals, all does. In July, 1924, a young buck, captured in 1923, in northwestern Nevada, was added to the herd, raising the number to 7. TEXAS Formerly antelope abounded on the plains of western Texas, but with the occupation of the country they have decreased until it has been possible to ob- tain definite information of only 42 existing bands, numbering about 2,400 animals, for the entire State. My principal sources of information concern- ing the antelope in Texas have been W. W. Boyd, State game, fish, and oyster commissioner, and C. R. Landon, in charge of the predatory-animal work of the Biological Survey in that State. On May 1, 1922, Mr. Boyd wrote that the antelope in Texas were ranging so far as possible in the rougher or sandy lands, owing to their having been hunted in high-powered automobiles. He added that one ranchman in 1922 reported 75 antelope fawns in the herd on his place the preceding year, and that he expected another good fawn crop that spring. Mr. Boyd is taking an active interest in the remaining antelope in the State and believes that the number can be materially increased. In December, 1924, through his deputies and other sources of information he completed the most thorough census of the surviving antelope in the State that has ever been obtained. The number proves to be much greater than was anticipated. One of Mr. Boyd’s deputies, Pete Crawford, writing on January 12, 1925, stated that a small herd of 4 or 5 antelope which ranged a few miles north of Marathon was completely wiped out. Mr. Crawford added that all the ante- lope herds that he has mentioned particularly in his report are protected by the ranchmen and popular sentiment. He stated that each of the herds that he personally knows is decidedly on the increase, and he believed that at the end of 10 years, if the present program of conservation is carried on, antelope herds in the region west of the Pecos will become a common sight. In the Houston Chronicle for November 12," 1923, it is stated from Hebron- ville that — “ Jim Hogg County perhaps can claim the only remaining antelope in south- west Texas. One herd of 16 to 20 ranges near town on the Hellen and Yeager ranches, while the other herd, somewhat larger, is on the W. W. Jones, Wilbur Allen, and Jonas Weil ranches in the southern part of the county. “ While protected by law and the ranchmen, as far as the latter are able to do so, yet occasionally one is killed by a hunter, as they are as gentle as range cattle and easily shot. “ The advent of the farmer in this section also is interfering with these beautiful animals and the time is not far distant when, like the buffalo which once roamed over these prairies, they, too, will have passed.” Apparently the antelope are on the increase in this district, since on Janu- ary 29, 1925, Mr. Boyd listed herds in that area totaling 2S5 animals, as indi- cated in area No. 40 on the map (fig. 16). H. G. Clark, of Lobo, Tex., writes that coyotes and eagles destroy some of the young fawns, and other causes contribute materially to reduce the in- crease. Mr. Landon on April 15, 1922, wrote that 5 or 6 years before he saw be- tween 30 and 40 antelope in one herd ranging near Big Lake, but that last fall the same herd contained only 7 animals. He added : STATUS OF THE PRONG-HORNED ANTELOPE, 1922-1924 53 “ There may be 4 or 5 antelope left on the Bar-S Ranch, north of Barnhart, but I am not certain that even this number survives. A few may exist on the 7-D Ranch or adjoining ranches near Stiles, but I doubt if there are at present 30 antelope in Reagan, Crockett, and Upton Counties, where they were formerly in great numbers. A small herd has existed on the Door-Key Ranch, 20 miles south of San Angelo, as long as I can remember, but I now understand there are only two left. These I saw about a year ago. “ On the McIntyre ranch, north of Sterling, there are possibly 20 or 30. Mr. McIntyre protects them as well as he can, but when they get outside of his fence they are usually killed. Sterling and the adjoining counties in all directions except to the east were formerly ideal antelope ranges, but the McIntyre herd is the only one now remaining in that section of the country, and I have been over practically every road in it. “ In the Panhandle a few herds of antelope still remain. On the holdings of the Matador Cattle Co., near Vega, I understand there are two or three small herds. One or more herds are also to be found on the holdings of Lee Bivans, of Amarillo. “ I believe it is conservative to state that where there were 10 antelope in Texas 10 years ago there is now less than one. In the country near Big Lake they suffered the greatest loss one fall three or four years ago, when through an oversight the legal protection of antelope in Texas was allowed to lapse for about 30 days. During this open season they were run down by men in motor cars and hunted so closely that the herd was practically exterminated. On the ranches near Amarillo, which were mainly posted, antelope would prob- ably have held their own so far as human agencies were concerned, but the ex- tremely severe winter of 1918, when the snow remained on the ground for weeks at a time, killed them by hundreds, and only a handful survive.” In a letter dated June 17, 1923, Charles Goodnight wrote that at one time he captured 5 antelope and placed them in one of his fenced pastures near Good- night, where they increased to 18, after which all were killed by hunters. He added that in his opinion antelope will not live in small incl.osures but do well where they have plenty of room. In the following summary of existing antelope in Texas all the reports are as of December, 1924, unless otherwise stated (fig. 16) : 1. Band of 6 reported in Hansford County. 2. Band of 3 reported in Ochiltree County. 3. In 1922, 25 antelope were located on the Sheldon range in Oldham and Hartley Counties, and a band of 4 on the Bivins ranch, near Cbanning, Hart- ley County. In December, 1924, the State game department estimated 350 for the entire county. 4. In 1923 there was a band of 5 in the breaks of Moore County ; these are now estimated to number 25. 5. A band of 16 is reported from Hutchinson County, apparently about main- taining its numbers. 6. Forty were reported in 1924 in Roberts County. 7. In 1922 a band of 7 were living on the Landergin West ranch, near Adrian, Oldham County; in December, 1924, 125 antelope were estimated to exist in this county. 8. In 1922 60 were reported in the Brown and Trujillo pastures, near Amarillo, Potter County; in December, 1924, only 20 were estimated to be found in this county. 9. Twenty-five are estimated to be living in Carson County. 10. Deaf Smith County is estimated to contain 100 antelope. 11. A band of 10 is reported in Randall County. 12. Fifteen were reported in Castro County in 1924. 13. In 1922 a band of 15 was reported on the Francis Miller ranch in Bailey County; in 1924 a total of 50 was estimated in this county. 54 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE 14. A band of 9 was reported in 1923 on Spring Lake ranch, Lamb County ; in December, 1924, 30 were estimated in this county. 15. One hundred antelope were estimated in the fall of 1924 in Cochran County. 16. Seventy-five are reported in Hockley County. 17. A band of 10 is reported in Lubbock County. 18. Five antelope are reported in Kent County. 19. In the spring of 1924 a band of 5 was reported about 18 miles west of Seminole, and another band of the same number about 16 miles west of Sea- graves, Gaines County ; in the fall of that year the total number in the county •was estimated at 30. 20. A band of five is reported in Borden County. 21. A band of five is reported in Scurry County. •DALLAM f 5 UPS. HAPJUZ (350) (If r T mzz. (Tof)" TT OOHLtl Flo. 16. — Distribution of antelope in Texas, estimated at 2,407, in 43 areas 22. Twenty-five are reported in Andrews County. 23. Fifteen are reported in Mitchell County. 24. In 1922, 10 to 12 were reported as ranging in northern Hudspeth County, southwest of Orange, N. Mex. 25. Several small bands, aggregating about 65 antelope, were reported in 1922 as occurring in central Hudspeth County. A. E. Gray, in charge of the rodent work of the Biological Survey in Texas, writes that the consensus of opinion among ranchmen is that these animals are gradually decreasing. The stockmen state that, in addition to the antelope killed by local residents, numbers are killed by hunters from El Paso who make annual trips into that country during the deer season. In the fall of 1924, however, the State game department estimated a total of 125 antelope in Hudspeth County ; in the absence of details as to their distribution it may be assumed that this indi- cates an increase in herd No. 25. STATUS OP THE PEONGLIOENED ANTELOPE, 1922-1924 55 26. In 1922, 75 were reported on the W. D. Casey ranch in northeastern Culberson County. 27. A band of six was reported in Loving County in 1924. 28. Fifty were reported in Midland County in 1922, but the present status of the band is not known. 29. Fifteen antelope were reported in Glasscock County in 1924. 30. In 1922 about 30 were living on the McIntyre ranch, north of Sterling ; in the fall of 1924, 125 were estimated in Sterling County. 31. A band of 7 was reported in 1923 near Big Lake, Reagan County ; in the fall of 1924, 14 were reported in this county — probably the same band. 32. Fifty were reported in 1924 in Irion County. 33. In 1922, 13 antelope were reported six miles north of Valentine, and S on the Jones Ranch, in northwestern Jeff Davis County; no later report concerning these has been received. In 1924 reports give 21 in the western part of the county. 34. A herd of 20 was reported in January, 1925, ranging on a ranch five miles southwest of Fort Davis. About 75 head were reported also, scattered in small herds, on the H. L. Kokenot ranch along the border line between Jeff Davis and Brewster Counties. 35. In 1922 several small bands were reported east and south of Fort Stock- ton, Pecos County, and in 1924 about 20 antelope were reported in this area. 36. About 300 antelope were reported in 1924 to be living on the Fisher ranch, 17 miles southwest of Marfa, in Presidio County ; these animals are rigidly protected by the owner. On the Cardwell ranch-, seven miles west of Marfa, there is a herd of about IS head. 37. A herd of about 25 is located about three miles northeast of Alpine, Brewster County, in one of H. K. Kokenot’s pastures, the animals being care- fully protected by the owner. A band of four or five ranges near Altuda in the same county. 38. A band of eight antelope is known to range between Dryden, Terrell County, and the Rio Grande, the band having increased from three animals in 1922. 39. Thirty antelope were reported in the fall of 1924 in Webb County, but the exact locality was not stated. 40. In January, 1925, State Deputy Game Warden O. R. Stephens reported several bands, aggregating about 285 antelope, as ranging mainly in Jim Hogg and Zapata Counties, as follows : “ On the W. H. Yager ranch, situated in the corners of Jim Hogg and Webb Counties, there is a band of antelope numbering 35 head, 17 of which were counted as fawns the past summer. On the W. W. Jones ranch, located in the eastern part of Jim Hogg County, there is a band of about 150 head, 40 of which were counted as fawns last summer. Another band of 40 lives on the Jonas Wiles ranch in the southeast corner of Jim Hogg County, and another of 60 head on the Wilbur Allen ranch in the south part of Jim Hogg County.” 41. Forty antelope are reported from Brooks County. 42. Fifteen are reported by the State game department from Hildago County. UTAH Antelope were once plentiful and widely distributed over the greater part of Utah. Gradually they have been reduced in numbers until now we have been able to learn of survivors existing in 10 sections of the State, numbering about 670 animals. The information has been mainly obtained through the efforts of George E. Ilolman and B. B. Richards, in charge, respectively, of the predatory-animal and rodent-control work of the Biological Survey in the State, with the assistance of D. II. Madsen, State game warden. Owing to the size of the State and to the fact that the surviving animals occur mainly in the sparsely settled districts, it has been exceedingly difficult to gather ac- curate information as to the exact number of these animals, but it is believed that there are few additional to those here reported. Mr. Madson has ex- pressed interest in the conservation of the antelope, and on December 8, 1923, wrote : 56 BULLETIN" 1346, U. S. DEPARTMENT OF AGRICULTURE “ During the past few months we have interested the Union Pacific Railroad Co., which is at present working out plans for the development of the Zion Park and Bruce Canyon scenic attractions, and which has agreed to give us all the cooperation possible in the protection of the antelope.” The information appears to indicate that the antelope in Utah are rapidly decreasing. Very definite and prompt efforts will be necessary to prevent their complete extermination. It is to be hoped that local game-protective associa- tions and others will make special efforts to safeguard the few widely scattered surviving bands. It is gratifying to note that in certain areas, as on the Escalante Desert, in Iron County, the settlers are interesting themselves in antelope protection. Unfortunately, reports from remote districts indicate that herders make a practice of killing these animals when opportunity offers. The antelope in Utah are distributed as follows (fig. 17) : 1. In 1922 a band of about 50 was ranging in the vicinity of Erickson, in Tooele County, where they were reported as being killed, especially in winter, and in danger of extermi- nation. Another band of 30 was ranging from Callao, in Juab County, to Gold Hill, in the same county. 2. A band of 50 was re- ported in 1922 as ranging in the vicinity of Cherry Creek, in Juab County, where they were said to be maintaining their numbers. 3. In 1922 a band of 50 was reported in Snake Val- ley, Millard County. These are the survivors of the herd of about 200 there in 1919. Their decrease is attributed both to their be- Fig. 17. — Distribution of antelope in Utah, estimated ing hunted and to the in- at 670, in 10 areas roads of predatory animals. 4. In 1922 a band of about 20 was reported in White Valley, Millard Count y. 5. In 1923 a band of not less than 75 was reported in the vicinity of Sevier Lake, in Millard County. This is said to be increasing. 6. Several bands aggregating about 150 animals were living in 1923 on the desert in Emery and Wayne Counties, ranging thence down to the Green River breaks. A few were reported on the east side of the Green River, in Grand County. 7. In 1922 about 50 were reported in the vicinity of Milford, in eastern Beaver County. 8. About 50 antelope are reported to live on the desert between Lund and Cedar City, in Iron County. Travelers on the road between these two places not uncommonly see some of these animals. As many as 50 have been seen on one trip. The settlers are interested in their protection and the antelope have become very tame. L. L. Carter, who has been long familiar with that region, states that in 1919 there were about 250 antelope there. After a period of heavy decrease it is believed that under the present protection they are now increasing. 9. About 100 antelope were reported in 1922 as about maintaining their numbei’S in Hamblin Valley, northwestern Washington County. Another band of 20 is reported from Pine Valley, in the same county, concerning which STATUS OF THE PRONG-HORNED ANTELOPE, 1922-1924 57 Mr. Carter states that predatory animals and shooting have caused a reduction from about 50 present in 1919. 10. In 1922 a band of 25 was reported in Hurricane Valley, Washington County. WYOMING Wyoming has the distinction of possessing the largest number of antelope surviving in any State. This, however, is only a pitiful remnant of the vast numbers which once roamed its great open plains. Antelope are now reported from 27 sections and total 7,000. In 18S5 on the Big Sandy River they were estimated to number about 30,000, or as many as now survive on the whole continent. Practically throughout the United States, as in Canada and Mexico, there is now a close season on antelope. A modified exception to this rule exists in Wyoming, where an act approved February 18, 1921, which still remains in force, reads as follows : “ Whenever, in the judgment of the State game and fish commission it is deemed desirable, the said commission may direct the State game and fish commissioner to issue not to exceed one hundred special buck antelope permits.” Owing to the numbers of antelope in some sections of Wyoming in 1922, plans were made for the issuance of 100 buck-antelope licenses under this law, but so strong was the public opposition which developed that the idea was abandoned. The history of the Greybull River herd on the Pitchfork Ranch and vicinity, above Meeteetse, is a good illustration of the manner in which an antelope herd may be built up and also demonstrates the fact that a great increase of game under protection in the midst of a cultivated district may become detrimental to the interests of the farmers and lead to open antagonism toward the ani- mals. The late L. J. Phelps, one of Wyoming’s pioneers, living at Meeteetse, many years ago realized that the antelope were disappearing. In 1902 he de- clared that no antelope should be molested anywhere on his holdings and pro- hibited shooting. At that time there was a band of about 15 ranging in the vicinity of the Pitchfork Ranch. Through Mr. Phelps’s influence during the next 21 years the original 15 increased to about 1,500. During 1923, Charles J. Bayer, in charge of the predatory-animal work of the Biological Survey in Wyoming, visited the Pitchfork Ranch to investigate this herd, and reported that there were practically 1,000 antelope ranging within the boundaries of the territory of area No. 3 on the accompanying map (fig. IS). They were broken into bands of from 25 to 125 each. It is planned to verify the numbers by a count during the fall of 1925. Eugene Phelps, in charge of the Pitchfork Ranch holdings, reports to Mr. Bayer that during the past two years the animals have increased to such an extent that they have become a pest. They enter grainfields after harvest and consume much grain before it can be hauled in and threshed ; they also graze throughout the year on lands owned and leased by his company. He contends that the antelope consume sufficient forage from their holdings to accommodate easily a good-sized band of sheep, and this con- tention appears to be correct. Many of the antelope were grazing in the pastures and fields at the time of this investigation. A. M. Hogg, representing the Ilogg brothers’ land holdings in that vicinity, reports that their company has suffered considerable loss of forage through the antelope. He states that during the past four years the antelope have cleaned all the forage from one field of 1G0 acres. Immediately after a heavy snowstorm on October 24, 1923, a band of between 500 and 700 crossed I he fence into an alfalfa and grain field and consumed about half the second crop of forage. They also visited the grainfields at night and destroyed some 58 BULLETIN 1346, U. S, DEPARTMENT OP AGRICULTURE grain and ate Lay in nnfenced stacks but without much damage. With the melting of the snow and the return of favorable weather three weeks later, the antelope left the ranches for the outside range. It is reliably reported that after a heavy snowstorm in October, 1922, about 500 antelope drifted out of the Meeteetse Valley into the farming country around Burlington, where the farmers shot large numbers, keeping the band moving in an easterly direction until they crossed the Big Horn River. It is impossible to learn how many animals were killed before they crossed the river, but everything- indicates that at least half of the band was hung up in meat houses along the way. As near as can be ascertained the survivors did not return. Eugene Phelps, A. M. Hogg, and Forest Supervisor Andrew Hutton suggest that a limited-license system should permit killing 50 to 100 buck antelope for each of the nest two years, in this way ridding the range of many old ani- mals. According to their statement there are twice the number of bucks really needed for the welfare of the herds, and this recommendation was made with the belief that such killing would cause large bands to split and spread into the adjoining areas, thus avoiding their congestion in one central district. If after a couple of years it should be found that the antelope have been properly distributed and are not in sufficient numbers to cause material damage to crops, then the season could be closed again. The census of antelope in Wyoming has been compiled mainly by Albert M. Day and Charles J. Bayer, of the Biological Survey, with the cooperation of Frank S. Smith, State game warden. The distribution of the antelope in the State is approximately as follows (fig. IS) : 1. The Yellowstone Park herd comprises the antelope which in summer fre- quent the plains of the upper Yellowstone River, within the boundaries of the Yellowstone Park. During severe winters, particularly when the snowfall is heavy, they have generally been forced to descend along the valley of the Yellowstone River to lower country in Montana. In 1909 the Yellowstone Park antelope herds were estimated to number about 2,000 animals. The last heavy loss occurred in the winter of 1921-22, when the deep snow made it difficult for them to escape the depredations of coyotes and wolves, and others perished from starvation. In the spring of 1922 only 235 remained. Horace M. Albright, superintendent of the Yellowstone National Park, to whom the writer is indebted for the information concerning this herd, in a letter dated September 10, 1924, stated that during the summer of 1923 approximately 70 fawns were born, of which all but 5 survived the following winter. In the spring of 1924 there were approximately 320 antelope in the herd. Reports from the summer range indicate that a large number of fawns were born, and in December, 1924, Mr. Albright reported 410 antelope in the park herd. He arranged to feed and safeguard the animals during the winter of 1924-25. The decrease in numbers in this herd through a series of years appears to have been brought to an end under Mr. Albright’s guardianship. 2. In this area about 80 antelope range along the Shoshone River, in Park County. 3. One of the largest single herds in the State is reported ranging on the Greybull River, in southern Park County. It is estimated to contain approxi- mately 1,000 animals and to be increasing. Further details concerning these are given above. 4. There are about 100 antelope southwest of Burlington, in Big Horn County. 5. Bands estimated to number 200 range in the Stagner and Black Moun- tains and on Owl Creek, Hot Springs County. 6. About 150 antelope are located near Kaycee, in Johnson County, where they are said to be decreasing rapidly. 7. A band of about 60 is reported on Wattel and Hanging Woman Creeks, in northeastern Sheridan County. STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 59 S. About 130 are reported about 7 miles southwest of Gillette, in Camp- bell County. These are said to be fast decreasing as a result of hunting. 9. In this area approximately 350 antelope range on the Belle Fourche River, in Campbell County. Their numbers are reported to be rapidly de- creasing through shooting. 10. About 60 antelope are reported along the Little Missouri River and the North Fork of the Cheyenne, in Crook County. These animals undoubtedly range back and forth across the border into 'Montana. They are reported to be rapidly decreasing. 11. One hundred and fifty antelope are reported to range on Lodge Pole, Prairie Dog, and Black Thunder Creeks and Cheyenne River in Weston County. 12. About 300 range along Antelope, Bear, and Sand Creeks, in northern Converse County. 13. Three bands, totaling about 70, appear to be generally scattered over the northern half of Niobrara County. 34. In southern Niobrara County is a band numbering about 180, and there are 90 near Raw Hide Butte, in northern Goshen County. These herds appear to be about holding their own. 15. This area, covering part of northern Platte County, is reported to have about 330 antelope, in two bands of 150 each, ranging on Glendo and Flat Top Creeks and Laramie River, and one band of 30 on Fish Creek. 16. I’li is area, in middle (-astern Platte County, is reported to have about 75 antelope, mainly about Goshen Hole and Deer Creek, near Wheatland. 17. In the middle eastern part of Albany and southern Platte C’ountieSi there are about 142 antelope, made up of three bands, numbering, respectively, 32, 45, and 85, ranging on Sibylee, Antelope, and North Chugwater Creeks. These are reported to be decreasing as the result of shooting. 18. About 150 antelope are reported as ranging on Horse and Bear Creeks, in Laramie County. These are said to be decreasing rapidly. 19. A band of 15 lives on Mule Creek in northern Albany County near Marshall. 60 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE 20. About 25 are reported in Shirley Basin, in northeastern Carbon County. 21. A band of 85 is reported in Natrona County, along Powder River and Bates, Poison Spider, and Fish Creeks. 22. This area contains about 400 antelope ranging on the Sweetwater Divide, in Freemont County. 23. One hundred antelope are reported on Big Sandy Creek and New Fork of Green River, in Sublette County. 24. About 150 are reported near Fontenelle, in Lincoln County. They are said to be decreasing rapidly through shooting. 25. About 75 occur on Muddy Creek, in Uinta County, where they are said to be rapidly decreasing. 26. In the Green River Valley, in western Sweetwater County, about 1,000 antelope are reported, which makes it the second largest herd, and gives Sweet- water County a total of 2,100 antelope (see area No. 27), by far the largest number surviving in any county in the State. 27. This area is reported to include about 1,100 antelope, the largest num- ber in any similar area in the State. It lies mainly in southeastern Sweet- water County, extending into the adjoining part of Carbon County. The main bands range on Black Rock, Shell, Skull, and Lost Creeks. CANADA In Canada antelope are now limited to the Provinces of Alberta and Sas- katchewan. In Alberta bands are located in 5 areas, containing a total of about 1,030 animals. In Saskatchewan they are located in 9 areas in which are about 297 animals, or a total of 1,327 north of the United States. Antelope formerly ranged east into Manitoba and north to the limit of the plains along the Saskatchewan River. For some 3rears the antelope in Canada diminished rapidly, but are reported now to be about maintaining their numbers or even increasing in some areas. Although the conservation of antelope in Canada is mainly a matter for the attention of the Provinces, yet the Dominion Gov- ernment assumes general responsibility in regard to all wild life, particularly concerning the antelope. The Canadian National Park at Nemiskam was especially created for their protection. J. B. Harkin, commissioner of Canadian national parks, states: “ The question of creating other parks as sanctuaries is now receiving the attention of the department. Our efforts are being retarded, however, owing to lack of appropriations necessary to proceed with the work. A limited num- ber of young antelope are being raised on the national antelope refuge at Nemiskam and will be transferred to Buffalo National Park at Wainwright, Alberta, as soon as they are old enough to be shipped. These, with the antelope already at the park, will form the nucleus of a new herd. I think it can be safely said that, due to the efforts put forth in recent years for the protection of these animals, they have not seriously decreased and are now holding . their own.” Concerning the heavy losses of antelope which were reported to have taken place by their drifting against fences along the railroad in this region some j^ears ago, F. Bradshaw, game commissioner of Saskatchewan, writes that no serious recent losses of this kind have occurred, but that he read an article not long ago by Doctor Prince in Rod and Gun in Canada, in which reference was made to thousands of animals dying along the fence of the Canadian Pacific Railway west of Swift Current. For the information concerning the surviving antelope in Canada the writer is indebted to J. B. Harkin, commissioner of Canadian national parks ; to F. Bradshaw, game commissioner of Saskatchewan ; and to Benjamin Lawton, chief game warden of Alberta. The distribution of the herds is as follows (fig. 19) : 61 STATUS OF THE PRONGHORNED ANTELOPE, 1922-1924 ALBERTA 1. About 500 are reported on the north side of Bow River above its junction with Lethbridge or Belly River, west and south of Brooks, on the Canadian Pacific Railroad. 2. About 100 range on Red Deer and South Saskatchewan Rivers, a short distance west of their junction. 3. About 100 are reported in the section between Belly River and Bow River, to the northeast of Lethbridge. 4. On the National Antelope Refuge in Nemiskam, to the west of Lake Pakowski, ISO antelope were reported in September, 1924. 5. The latest information, in 1924, gives about 150 as ranging in the extreme southeastern corner of the Province. SASKATCHEWAN 6. About 40 antelope are said to range on both sides of the South Saskatche- wan River, west of Owensville. 7. To the northeast of White Bear Lake about 20 antelope are said to range. 8. A band of 8 antelope is reported on the South Saskatchewan River a few miles west of Saskatche- wan Landing. 9. Between White Bear and Luck Lakes, some distance north of the South Saskatchewan, 12 ante- lope are reported. 10. A band of about 10 are reported near Long Valley, northwest of Lake Chaplin. 11. About 100 range about Bigstic-k Lake, north of Maple Creek on the Canadian Pacific Railroad. 12. To the north of Cy- press Lake, in the south- western corner of the Province, 40 antelope are reported. 13. On the north side of Frenchman Creek, near the town of East End, 27 are reported. 14. About 40 are re- ported to occur in the area south of Wood Mountain, drained by Frenchman Creek and Poplar River, both tributary to the Missouri. MEXICO It has not been possible to obtain definite information concerning the dis- tribution of the antelope bands or the numbers contained in them from any part of Mexico except Sonora. The accompanying maps (figs. 20 and 21) and statements concerning the surviving antelope in that country are based on personal knowledge of the writer and on information mainly received from Carlos Lopez, in charge of the Federal game administration of Mexico, and from Game Warden Ben Tinker. Formerly antelope ranged south over the great Mexican tableland to within less than 100 miles of tin; City of Mexico. It is interesting to know, as set forth earlier in this report, that the first mention of antelope seen by Europeans Fig. 19. — Distribution of antelope in Canada, estimated at 1,030 in 5 areas in Alberta, and 297 in 9 areas in Saskatchewan ; a total of 1,327 in 14 areas 62 BULLETIN 1346, U. S. DEPARTMENT OE AGRICULTURE on this continent was recorded in an account of a great hunt organized for the viceroy of Mexico in 1540, at a point near the present station of Cazadero on the Mexican Central Railroad in extreme southwestern Hidalgo. The main herds of antelope in Mexico are undoubtedly located on the broad arid plains of Coahuila, Chihuahua, and northeastern Durango. Other herds occupy considerable territory in northwestern Sonora, some occasion- ally ranging back and forth across the border between Sonora and Arizona, and others ai’e located in Lower California. From information received it appears possible at this time that, in general, antelope may be holding their own in Mexico. On October 1, 1922, a close season of 10 years on antelope, which had been established by President Obregon, became effective. This should serve to lessen the number of these animals killed and so favor their increase. There are great areas of sparsely occupied plains on the northern Mexican table- lands where they might find a home far into the distant future. • To assist in safeguard- ing the antelope and other game animals of northern Sonora the Permanent Wild Life Protection Fund, through Doctor Horna- day, has entered into an agreement with the Mexi- can Government whereby it employs Ben Tinker as game warden, with head- quarters at Tucson, Ariz., to work along both sides of the Arizona-Sonora bor- der to prevent poaching. It is conservatively esti- mated that there are 2,395 antelope in Mexico, of which 1,300 are estimated to be in Chihuahua, Durango, and Coahuila, and 500 in Lower California. The remainder are more definitely known in Sonora. These numbers will serve as a working basis until there is opportunity to get more complete information. It is probable that there may be many more on the plains of Chihuahua and Coahuila than here estimated. The following details of distribution are based on a letter received from Professor Lopez in January, 1924, and from other information available on the subject. It is grouped under States and the Territory of Lower California, as follows (figs. 20 and 21) : COAHUILA Fig. 20. — Distribution of antelope in parts of Mexico — - in Chihuahua, CoaLuila, and Durango, estimated at about 1,300 in 2 areas (see also fig. 21) In the great Yalley of La Encantada, to the west of Muzquiz, bands of 50 to 100 antelope occur. They are also about the Hacienda de San Antonio, and are more abundant about the Hacienda de Paila and on the plains about the neighboring mountain range of Espianzo. Most of the antelope in Coa- huila are located west of the railroad which runs south from Eagle Pass, Tex., to Saltillo, and north of the railroad extending from the last-named place westward to Torreon. STATUS OF THE PRO UGH OR NED ANTELOPE, 1922-1924: 63 CHIHUAHUA 111 extreme northwestern, southeastern, and eastern Chihuahua antelope occur in varied numbers. The bands in the extreme northwestern part are separated from those which range along the Mexican Central Railroad to the east. The great Bolson de Mapimi and the region east of the Mexican Central Railroad is a vest territory ideally suited to the needs of these animals. In southern Chihuahua antelope occur on both sides of the Mexican Central Rail- road, particularly along the border of Durango. DURANGO In Durango antelope are now limited mainly to the northeastern part of the State, in the district of San Dimas. Antelope are reported to be rather com- mon in the following local- ities : Lapioriz, Mara- velles, El Pilar, Santa Rita, San Julian, Las La- gunas, Huachinepas, San Francisco de los Lobos, Pericos, and Huahiapa y Gavilanes. They are also said to be abundant in the district about Escalon, along the border of Chi- huahua and Durango, near the base of the Sierra del Diablo. Antelope in Sonora are practically all west of the railroad extending from Nogales on the Arizona border south to Guaymas and in the region lying north of a lind drawn from Hermosillo west to the coast of the Gulf of Cali- fornia. A few bands in northwestern Sonora Fig. 21. — Distribution of antelope in Lower California and Sonora, Mexico, estimated at 500 in 2 areas of Lower California, and 595 in 4 areas of Sonora ; a total of 1,095 in 6 areas (see also fig. 20) range back and forth across the Arizona border. It is these bands which are now under the guardianship of the Permanent Wild Life Protection Fund. Under date of January 4, 1925, Ben Tinker, who represents the Permanent Wild Life Protection Fund along the Sonora-Arizona border, supplied the writer with interesting information concerning the distribution of the sur- viving antelope in Sonora. They are reported to occupy 4 areas and to have totaled 595 animals in November, 1924, when they were counted by him. Fol- lowing is his summary of these antelope herds: 1. Comprises numerous bands, numbering 459 all told, ranging from the southern end of the Sierra Rosario south and east to the Sierra Blanca and the Rio Sonoyta, thence eastward (north of Sierra Pinta) to the eastern side of the Sierra de San Francisco. The largest single band, containing 73 ani- mals, ranges between the Sonoyta River and Sierra de San Francisco during the months of October, November, and December and southward from this river to the Sierra Pinta during the remainder of the year, 64 BULLETIN 1346, U. S. DEPARTMENT OF AGRICULTURE 2. Comprises 17 in two bands between the Sierra de la Nariz and the town of Altar. 3. Comprises 56 in small, scattered bands from Sierra del Cajon eastward to within six miles of Noria Station on the S. P. de M. Railway. 4. Comprises 63 in many small bands between the Rio San Ignacio and the city of Hermosillo. LOWER CALIFORNIA Antelope in Lower California are distributed mainly on the plains east of the central mountain range from the California border south to the middle of the peninsula. They are also on the desert of Vizcaino, where they live west of the main mountain range, reaching the borders of the Pacific on the shores of Vizcaino Bay on the north and Ballenas Bay on the south. It is estimated that not less than 500 antelope survive on the peninsula. Formerly antelope in Lower California ranged south beyond Magdalena Bay, but for many years they have been extinct over a large part of their former territory. During the past 15 years antelope have been continuously hunted in Lower California, and it is rather surprising that they have continued to survive. It is hoped that the operation of the present close season on them may result in their numbers again increasing. Natural conditions are such that Lower California will never be densely populated or occupied by farming communities of any im- portance. Water is scarce in the interior, and great plains covered with desert vegetation afford an ideal home for antelope. With reasonable protec- tion they might survive there in large numbers far into the future (fig. 21). ADDITIONAL COPIES OF THIS PUBLICATION MAT BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 15 CENTS PER COPY V . UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN No. 1349 Washington, D. C. ▼ September, 1925 THE BROOD-REARING CYCLE OF THE HONEYBEE By W. J. NOLAN, Associate Apiculturist Division of Bee Culture Investigations, Bureau of Entomology CONTENTS Page Introduction 1 Method 4 Annual Brood-Rearing Cycle 6 Description of the Colonies Used in 1921 12 Seasonal Characteristics of 1921 13 Brood Rearing of a Typical Colony for Two Successive Seasons '. . 14 General Observations on the Remaining Colonies 16 General Discussion of the Records for 1921 25 Observations in 1920 27 Migrations of the Queen Within the Hive 30 Compactness of Brood Nest 32 Time Relation of Brood Rearing to Nectar Gathered . .' 33 Egg Laying 35 Conclusions 36 Literature Cited 37 Tables 38 Graphs 44 WASHINGTON GOVERNMENT PRINTINC OFFICE 192S ’ I UNITED STATES DEPARTMENT OF AGRICULTURE DEPARTMENT BULLETIN N®. 1349 Washington, D. C. ^ September, 1925 THE BROOD-REARING CYCLE OF THE HONEYBEE By W. J. Nolan, Associate Apiculturist , Division of Bee Culture Investigations, Bureau of Entomolgy CONTENTS Page Introduction 1 Method 4 Annual brood-rearing cycle 6 Description of the colonies used in 1S21 12 Seasonal characteristics of 1921 13 Brood rearing of a typical colony for two suc- cessive seasons 14 General observations on the remaining colonies, 16 General discussion of the records for 1921 25 Observations in 1920 27 Page Migrations of the queen within the hive 30 Compactness of brood nest 32 Time relation of brood rearing to nectar gath- ered 33 Egg laying 35 Conclusions 36 Literature cited 37 Tables 38 Graphs 44 INTRODUCTION In previous work of the Bureau of Entomology emphasis has been placed on the conditions necessary for the proper wintering of bees, in order that colony population and energy may be conserved to the utmost during the period when no brood is reared by normal colonies. It is evident, however, that merely wintering the bees in the best possible condition will not in itself guarantee that the colony will at the right moment have the proper strength and composition for gathering a maximum, honey crop. Nevertheless, if, through proper wintering, the strength of the colony has been adequately conserved, the resumption of brood rearing in the spring may take place at the proper time and the amount of brood reared may increase at a remark- able rate, since the ability of the colony will not have been impaired through excessive work during the winter. It is appropriate, there- fore, that the investigation of wintering conditions should be followed by an investigation of the factors which modify brood-rearing activity, more especially those which are under the control of the beekeeper. Since normally a worker bee, before going to the field, spends the first two or three weeks of its life in duties within the hive, the quantity of nectar gathered by any colony depends not merely on the total number of bees in the colony during ahoncy flow, but on the number included within that total which represents bees of proper age to serve as nectar gatherers. In order to have the largest possible number of field bees at the proper moment, therefore, the highest daily rate of bees emerging from the brood cells during any given season should 46969° — 25t — Bull. 1349 •1 2 BULLETIN" 1349, U. S. DEPARTMENT OF AGRICULTURE occur about three weeks in advance of the main honey flow; in other words, the queen should reach her maximum daily egg-laying rate during the period six weeks prior to the honey flow. Since in a colony left to itself such is usually not the case, a correct understanding of the principles governing brood rearing throughout the year becomes of prime importance to the beekeeper, if he is to handle his colonies in such a way as to secure a maximum honey crop. Lack of knowledge of the principles governing brood rearing may cause a reduction in the honey crop by bringing about in a colony any or all of the three following possibilities : 1. The population of the colony may not become large enough to provide sufficient field bees during nectar flows to gather surplus adequate to give the beekeeper a fair return for time spent and capital invested. 2. Surplus honey may be consumed in regions of early nectar flows by bees which have emerged too late to serve as nectar gatherers, and too early to winter over or even to assist in building up the colony for winter. 3. Swarming may be stimulated if the ratio between hive bees and field bees does not remain such as will avoid causing a congestion within the hive whenever one of these classes is relatively idle while the other is extremely busy. The prevention of any or all of these states involves such questions as wintering, stores for spring, requeening, population of the colony at the beginning of brood rearing, swarm control, dequeening, removal of brood, and other related factors. In short, regardless of its immedi- ate purpose, every sound beekeeping practice having to do with the actual manipulation of the colony itself has as its final result the eli- mination or prevention of some one of the three above-mentioned conditions. The utility of any manipulation of the colony may well be gauged by the extent to which such an outcome is achieved. It is essential, then, to have a clear understanding of the principles of brood rearing in order to apply the proper procedure to any case so as to obtain the desired result. The manner of increase in a colony’s population has been under discussion since the days of the ancients. Views on this subject prior to the latter part of the seventeenth century, however, differed widely from those now held, since the sex of the queen had not yet been determined and many people even believed in the spontaneous generation or creation of bees. That brood rearing is a phenomenon in which the queen is concerned directly was not generally recognized until Swammerdam (14, p- 159 )1 in 1669 established clearly the actual relationship borne by the queen to any increase in the colony’s population. Since this great apicultural discovery, beekeeping litera- ture has been filled with reports and conjectures as to a queen’s daily egg-laying capacity, and the total amount of brood reared during a season. Among early investigators in the field, Reaumur (13, p. 475) in 1740 stated that the height of egg laying comes in the spring and that over a period of two months at that time the queen may average 200 eggs per day, this average being accepted for nearly a century afterwards as fairly typical of a queen’s egg- laying capacity. 1 Reference is made by number (italic) to “Literature cited,” p. 37 THE BROOD-BEARING CYCLE OE THE HONEYBEE 3 The first trustworthy determination of the number of eggs laid in a single day was made by von Berlepsch (3, pp. 68-69 ) in 1856. Having succeeded in confining the egg-laying activity of an especially prolific queen to a single comb for 24 hours, he found that mean- while 3,021 eggs had been laid. An estimate of the amount of brood remaining in the hive to which the queen belonged led to the assumption that she had been averaging nearly 3,000 eggs daily for the preceding 20 days. During the remainder of the nineteenth century this rate was widely accepted as a proper index of a queen’s daily egg-laying capacity, although von Berlepsch himself believed such a rate to be exceptional, and that a daily average of only 1,200 is probably usual. Inasmuch as this particular queen was active for five seasons, von Berlepsch assumed that she must have laid at least 1,300,000 eggs during her lifetime, a number which apparently has served many later writers as a basis for their estimates of the total possible egg-laying achievement of a queen. Baldridge (2) , an American contemporary of von Berlepsch, deserves mention because he furnished the first published census of all the eggs, larvae, and sealed brood in a modern hive, determined by an actual count. He even entertained the idea of counting all the eggs in a certain colony every 72 hours, but apparently never carried it into effect. The first authentic data as to the total number of eggs laid by a queen throughout an entire season were published by "Desborough in 1852 (8). In 1855 ( 9 ) he published data in regard to the number of eggs laid by one queen in two successive seasons, and in 1868 (10) he presented similar data covering six successive seasons for a single queen. Desborough’s figures were obtained by making periodic esti- mates of the area occupied by brood. The colony used seems to have been so much below normal strength, however, that his findings can not be taken as typical. For the next 40 years, of the many reports on the quantity of brood found in a hive, or of the daily egg-laying capacity of a queen, few are of any real value in understanding the annual brood-rearing cycle. Interesting as they may be, these reports too often represent only the performance of some exceptional queen during a single day at the height of the season. Such sporadic endeavors, either in themselves or in relation to other similar reports from localities under far different conditions, afford little basis for drawing conclusions as to brood-rearing activity throughout a whole season. Although during this long stretch of years it may have been realized that the annual brood-rearing cycle can be determined only by continuous observations on the same colonies during any given season, apparently no one undertook the task. Finally, in 1895, Baldensperger ( 1 ) furnished the first published results of successive counts or estimates throughout the year of the quantity of brood in a colony of normal strength. An epoch in this line of research is marked in 1901, when Dufour (11) published data obtained from the first comprehensive study of the subject by a scientific method of approach. As a result of four years’ work he had secured seven seasonal, brood- rearing records by actually counting, at intervals of approximately 21 days throughout each season, every egg, larva, and sealed brood cel), in each colony used. In 1912, the first seasonal curves based on results from brood-rearing investigations were presented by Briin- nich (4). In 1919 (5), and again in 1922 (6), he presented other 4 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE similar curves, representing the daily egg-laying rate of various individual queens. Briinnich’s work, unlike Dufour’s, is based, not on an actual count of each cell containing brood, but on a mathe- matical calculation of the number of such cells derived from linear measurements of the brood area on each frame throughout the season. From data thus obtained daily egg-laying rates are cal- culated for the whole season. Although the claim can not be made that Briinnich’s work is as nearly accurate as Dufour’s, the Swiss investigator has adopted a method which is fairly speedy and readily utilized, and which gives results reliable enough for most purposes. METHOD In 1920 work on this problem at the Bee Culture Laboratory was first begun when Lloyd K. Watson, formerly apicultural assistant, made actual counts weekly of all eggs, larvae, and sealed brood in five colonies for the entire season. Any such method of counting brood on each comb is necessarily slow. In cool weather it involves the possibility of brood becoming chilled before the operation is completed; at other times there is danger of robbing, and in any event there is too long a disturbance of the colony. Accordingly, when the writer took over the work at the beginning of the season of 1921, a photographic method was determined upon, whereby photo- graphs are taken weekly of every frame containing sealed brood, and counts are made later from the negatives. Only sealed brood is counted, because of its greater clearness on the negatives. As a result of the use of this method, photographic records of 16 colonies were obtained in 1921, and of 32 colonies in 1922. Adding to these the counts from the five colonies in 1920, the equivalent of a total of 53 individual seasonal brood-rearing records has been obtained already from the work now in progress. A small building adjacent to the apiary not only houses the camera permanently but also affords protection from robber bees while taking the pictures. During exposures two 500-watt electric lamps furnish light sufficient to obtain good negatives at all times within the building, regardless of conditions of light outdoors. The camera itself is fastened securely to one end of a base made of 2-inch plank. To maintain the brood frames firmly in position during exposures and yet to have in the negative an image of every cell on the exposed side of each comb, a substantial holder (Plate I, A) is used which consists of a base with two uprights at each end, the uprights being j oined by a top piece. The width of the holder is such that the lower half of each end bar of a Langstroth frame just fits into a groove extend- ing upward from the base on the inner surface of each upright. A super spring fastened to the rear edge of each groove presses the end bar firmly against the front edge, and thu§ the brood frame is held rigidly in a definite position, although it may easily be slipped in and out of the holder. The holder itself is fastened securely to the same base as is the camera, but at such a distance from the lens as to give a reduction to a scale two-thirds that of the original. Because of the uniform focal distance and the uniform illumination, all negatives are made on an identical scale and under the same light conditions. By the aid of a suitable device attached to the frame holder there is photographed with each frame of brood a record showing the date, the hive and hive body from which the frame came, Bui. 1349, U. S. Dept, of Agriculture Plate I Photographic Production of Brood Records A. — Apparatus used in obtaining records, showing camera, brood-frame holder with brood frame in position, wire net, record, and electric lamps. B. — Print made from a photographic record, illustrating the character of a permanent record THE BROOD-REARING CYCLE OF THE HONEYBEE 5 the location of the frame in the hive body, and the particular side of the frame. A net of wires forming 1-inch squares is permanently fixed to the holder at such a point as barely, to clear the surface of any brood frame in the holder and still be in focus. The squares, showing clearly in the negative (Plate I, B), divide the brood area into such small sections as to render possible an extremely accurate, direct count. For recording the counts from each negative, a card is ruled into squares identical in size and number with those in the picture itself. All squares corresponding to areas containing only sealed worker cells may be credited with the number of cells contained in 1 square inch; but in squares containing unsealed as well as sealed cells, the number of unsealed cells must be deducted first. There are many contradictory statements as to the number of cells per square inch, due in part to attemps to derive it mathematically from the dimensions of some single cell instead of counting the actual numbers in areas large enough to get a trustworthy average. Watson, by making such counts, found the average number per square inch to be slightly in excess of 26. This number has been used in the results presented here. Some variation exists between individual combs, however, possibly due to the foundation used. Thus, in combs from certain foundation the WTiter has found 26.3 worker cells per inch, and in combs from foundation bought after the results for 1921 were ob- tained he has found approximately 27 worker ceils per square inch. Much of this latter type of foundation has been used since that date, and subsequent results are therefore being calculated on this basis. It is very evident that the general relations of a curve based on the amount of brood counted will remain the same, regardless of whether in the counts the number of cells per square inch is taken to be 26, 27, or some other figure of nearly the same size. As a matter of fact, the difference between 26 and 27 is less than 4 per cent, or less than 40 in every 1,000. In areas containing 50 per cent or more of unsealed cells it has been found preferable to count each individual sealed cell. Sealed drone cells also are counted individually. They have little influence on the totals, however, because by proper selection of brood combs it has been possible to keep the total of sealed drone cells well below 600 on any one count. Since in counting individual cells, either sealed or unsealed, some arbitrary rule must be followed in crediting them to a particular square; all such cells are credited to, or deducted from, the total of the square immediately to the left of the vertical dividing line, or below the horizontal dividing line, as the case may be. Totals for each card and colony are calculated on an adding machine. The photographic record once obtained, the actual count- ing may be delayed until any convenient time, so long as the nega- tive docs not deteriorate. A series of such records permits making a year-by-year comparison of any portion of the sealed-brood area. Cut films, 5 by 7 inches in size, are used exclusively in this work. Before photographing, all of the frames containing sealed brood in any given hive body are shaken or brushed free of adhering bees, placed in an empty hive body, and immediately carried to the build- ing where the photographs are made. The exposures can be made in 10 seconds, thus keeping the frames out of the hive for an exceed- ingly short time. Danger of chilling the brood is thus reduced to a 6 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE minimum, although in any event, in the cooler weather of spring or fall, there are few frames containing sealed brood. The hive being open for so little time, danger from robbing is also reduced, and the normal activity of the colony is disturbed to a far less degree than is the case in any other method of brood-area determination so far employed. The great speed with which the work can be accomplished adapts this method to investigations on a large scale. In the first two years of this work more than 18,000 pictures were thus taken. DEFINITIONS The term “ brood nest ” as used in this bulletin applies to the space occupied by brood, regardless of the number of hive bodies in which brood is found. The term “ brood area” is similarly applied. The term “ super” is' used for the hive bodies which are placed above the second hive body to give additional room for colony activity, the lower two hive bodies being those which remain with the colony permanently, summer and winter. The hive bodies are of uniform size, regardless of purpose. Brood could be reared in the supers of hives in which no restriction was placed on the movements of the queen by a queen excluder. In the apiary of the Bureau of En- tomology, where this work was done, the colonies are arranged in groups of four for convenience in putting them into packing cases for winter, and such groups are referred to as packing-case groups. The term “quadruple packing case” refers, of course, to the fact that each of the packing cases used is capable of containing four colonies. “Nectar flow” and “honey flow” are used synonymously to cover those periods in which nectar available for the honeybee is secreted freely. “Pollen yield” refers to the gathering of pollen in large quantities. The term “natural requeening” is used in this bulletin for the requeening of colonies in which it is impossible to determine whether the old queen was lost through natural superse- dure or whether she was accidentally killed while the colony was being handled. Where artificial requeening was practiced, as by the killing of the old queen and the giving of a queen or queen cell, the period of queenlessness is less than in natural requeening. ANNUAL BROOD-REARING CYCLE Common recognition of certain factors underlying brood rearing has given rise to different apiary practices. For example, it has long been believed, regardless of geographical conditions, that a honey flow greatly stimulates egg laying. This is attested by such apiary practices as dequeening during a honey flow, removing brood, and the like. Although apiary practice has contributed much to a knowledge of brood rearing, it has not as yet furnished a clear, definite understanding of all of the factors causing an increase or decrease of brood-rearing activity. For instance, it has not been established as a fact that brood-rearing activity increases in a uni- form and regular manner during the beginning of the active season, nor that irregularities may then occur. In short, each seasonal phase of brood rearing presents problems, not fully solved as yet, which are of vital importance in beekeeping practice. It is a matter of common apiary experience that during a certain portion of the year, depending on weather, nectar flows, and other THE BROOD-REARING CYCLE OP THE HONEYBEE 7 conducive factors, brood-rearing activity is prevalent, and that in the remaining portion of the year brood-rearing activity is suspended. The annual brood-rearing cycle may therefore be divided into two parts: (1) a period of seasonal activity and (2) a period of seasonal suspension. The period of seasonal brood-rearing activity, or the “ active season,” takes place, roughly speaking, during the summer; the period of seasonal suspension of brood rearing, here called the “inactive season,” occurs in winter. SEASONAL ACTIVITY At the end of the inactive season, marked normally by the first incoming nectar or pollen, brood rearing is resumed and may proceed to a certain maximum in the fore part of the active season at a rate which is often strikingly noticeable. Brood-rearing activity during the remainder of the active season, up to the period of final decline, varies widely with geographical location or climatic conditions. In some regions it is maintained at a uniformly high rate throughout; in other regions it is broken in continuity by an interval of partial suspension. In still other regions there is a gradation between these two extreme types of seasonal brood-rearing activity. The final decline is either abrupt or gradual, depending also on geographical location. Regardless of geographical location or climatic conditions, the period of seasonal activity may be divided into the three following phases: (1) A period of initial expansion, (2) a major period, (3) a period of final contraction. It is not always possible to draw a sharp line of demarcation between these three seasonal phases of brood- rearing activity, because external environment, such as weather or incoming nectar and pollen, often causes the end of one seasonal phase to become so merged with the beginning of the next that the initial influence of the succeeding phase can not easily be detected. PERIOD OF INITIAL EXPANSION The initial expansion covers that period of the active season immediately following the inactive season, in which brood-rearing activity is normally resumed and is continued in spite of conditions which if occurring later in the season would tend to check brood rearing. It should be pointed out that the beginning of brood rearing here discussed is that caused by the incoming of the first nectar or pollen and does not apply to abnormal brood rearing during the inactive season, which will be discussed later. At no other time in the year does the tendency to rear brood seem so persistent as during this initial phase, except possibly when a colony has just swarmed or when a queen is beginning to lay for the first time. Although the rate of increase in brood-rearing activity may be greatly accelerated during the period of initial seasonal expansion by incoming nectar or pollen, the fact that such an expansion continues after resumption of brood rearing, even with no incoming nectar or pollen, indicates that this phase is purely seasonal and needs only the approach of spring to cause its appearance. Since this expansion in the spring is a seasonal phenomenon, and is bound to occur, the colony which will gain the most rapidly in population in the spring is the ono possessing the largest number of factors favorable for brood rearing. It becomes readily apparent, 8 BULLETIN 1349, U. S. DEPARTMENT OE AGRICULTURE. then, how important it is for the beekeeper to do everything in his power to have conditions in the hive just right at the moment this phase begins, if his colony is to get a good start from the very begin- ning of the season. Such action on the part of the beekeeper is especially imperative in regions where the honey flow follows close upon the opening of the active season, because under such circum- stances little time is given the colony in which to build up, and such time as is granted must be used to best advantage. In localities with early honey flows a successful season is dependent largely on the number of bees reared in the period of initial expansion. Regardless of any direct bearing upon the honey crop, it is to the advantage of the beekeeper to make the most of the tendency toward greatly heightened brood-rearing activity during the period of initial expansion, merely for the sake of having his colonies strong enough to resist certain diseases successfully. A colony which has gained a maximum population during the initial expansion is in a far better position to ward off European foul brood than is one which increases only slowly in the spring. It is too often the case in regions where this disease is prevalent that the nectar flows are not so correlated with the period of initial expansion as to result in a strong enough population to enable a c’olony to overcome the presence of this disease. That strength of population will minimize the effects of diseases of adult bees also is shown by the comparatively slight loss occasioned by Nosema apis in strong colonies. Morgenthaler (12) has stated his belief that a good prolific queen is one of the greatest aids in overcoming the Isle of Wight disease. In other words, the colony which successfully withstands the disease must be in good condition and strong enough to discount the loss in the adult popu- lation. Among all the invaders of hive or colony itself, the wax moth furnishes probably the most commonly recognized example of the importance and utility of a strong colony population as a curb to the harmful activities of invading organisms. For a colony to be strong throughout an entire season, however, a maximum increase in population must have been made first of all during the initial expansion. The successful beekeeper supplies conditions which cause the largest possible increase of colony population during the initial expansion, not only for the sake of obtaining a large number of honey gatherers during this period, but also to provide sufficient bees to resist chance inroads upon the colony. This especially applies in regions where natural conditions between the period of initial expansion and the main honey flow are not conducive to a sufficient increase in population to keep the colony from faffing an easy prey to certain invading organisms. Among some of the important factors which are within the power of the beekeeper to provide are a prolific queen, sufficient bees wintered over to meet all brood-rearing requirements in the spring, sufficient worker brood cells, sufficient stores of good honey, and proper insulation. All of these are factors which must and can be provided in the manipulations in the latter part of the previous season if the beekeeper wishes to -take the utmost advantage of the natural tendency toward intense brood-rearing activity at the begin- ning of spring. Conditions within the hive making for brood rearing during the period of initial expansion may be likened to a charge of explosives set to go off at a certain moment in the spring, the time THE BROOD-RE ARIN G CYCLE OF THE HONEYBEE depending on weather conditions; in the one case the force of the resulting explosion is definitely predetermined by the quantity of the charge; in the other case the amount of brood reared is definitely predetermined by the provisions made in the preceding season to give the colony the conditions most favorable for this purpose. THE MAJOR PERIOD The major period of brood-rearing activity extends throughout the active season from the time when normally the initial seasonal tendency would cease to make itself felt until the beginning of the period of final contraction. It is the longest of the three phases of seasonal activity. The character and sequence of honey flows under different climatic conditions cause brood-rearing activity during this phase to vary widely in different regions. Throughout the world, on the whole, brood-rearing activity during the major period falls either into one of two extreme types or into a gradation between the two. During this period one of the extreme types is marked by a con- tinuous high rate of brood-rearing activity, while the other extreme type is marked by a pronounced slackening or series of slackenings in such activity. In the third or intermediate type there is neither a con- tinuous high rate of brood-rearing activity during the major period, nor under normal conditions is there at any time a complete suspen- sion. This intermediate type, however, does show a seasonal slack- ening at some time within the major period. A continuous high rate of brood-rearing activity during the major period occurs in regions with a long inactive season in winter, fol- lowed by a short active season, usually characterized by overlapping honey flows. Following the prolonged period of winter suspension in such a region, brood-rearing activity during the period of initial expansion attains with striking rapidity a high rate, which, unless checked by conditions within the hive, is maintained throughout the major period extending over practically the whole of a relatively short active season and then, owing to the proximity of the period of winter suspension, undergoes an abrupt contraction. Such a type of seasonal brood-rearing activity during the major period is typical of subarctic conditions. A more or less complete suspension of brood rearing, or series of such suspensions, is found in regions with a short, almost nonexistent, inactive season, followed by a long season of activity, usually charac- terized by one or more periods of drought during the hottest weather. In such a region the initial expansion in brood rearing does not prog- ress so rapidly as in regions with a long, inactive winter season, nor is the final contraction so abrupt. The major period, instead of being characterized by a uniformly continued high rate of brood-rearing activity, is characterized by a pronounced midsummer slackening, or series of slackenings, in brood-rearing activity, probably caused by an absence of incoming nectar and pollen during the periods of drought. This type of brood-rearing activity during the major period is typical of tropical conditions. Intermediate between the two types just described is that typo of seasonal brood-rearing activity which exhibits a more or less marked slackening at some time during the major period, but never a com- plete suspension under normal conditions. This type may bo found 46969°— 25f— Bull. 1349 2 10 BULLETIN 1349, U. S. DEPARTMENT OP AGRICULTURE in the middle and southern latitudes of the United States. Brood- rearing activity after this period of midseason slackening may equal in intensity and extent that which took place before the slackening. An example of this type is found in regions of the southern Appa- lachian Mountains, where there is an early honey flow from the tulip- tree, followed by a midsummer dearth of nectar, which in turn is succeeded by a later honey flow from sourwood. In other regions where the continuity of brood-rearing activity is broken by a mid- summer dearth of nectar brood rearing during that portion of the major period having the larger amount of nectar available is much more active than it is in the other portion. In the vicinity of Wash- ington, for example, brood rearing is much more actively carried on before this midsummer dearth than afterwards. The main honey flow from tulip tree comes early, and the only later nectar flow of any consequence comes in September, and it often happens that even this yields little surplus. In the typical buckwheat region of the United States conditions are reversed in that the main nectar flow comes after a period of decreased brood-rearing activity in the middle of the major period. Certain regions, which otherwise, because of lack of nectar from natural sources, would show a tendency toward a midseason slack- ening of brood rearing, have had this tendency overcome through other agencies, such as the production of honeydew in midseason or by the introduction of plants which secrete nectar under conditions or at times when the native plants do not. This has happened in the southeastern part of the United States through the introduction of cotton and sweet clover. Such circumstances tend to cause brood rearing to remain at a higher level during 'certain portions of the major period than normally would be the case. Although the maximum nectar-gathering activity of the season takes place in the major period, the amount of surplus stored is really determined to a large measure during the major period of the pre- ceding season. This is true because the amount of surplus obtained during any season depends on the number of field bees available during nectar flows, and the number of field bees, in turn, depends largely on brood-rearing activity during the initial seasonal expan- sion. Granting a prolific queen, an abundance of worker cells, sufficient stores, and proper insulation at the beginning of the initial expansion, however, the amount of brood reared during that period will depend largely on the number of bees reared during the latter part of the preceding major period and which have wintered over. It follows, then, that one of the most important aspects of the major period lies in the fact that during it the conditions arise which will lead to success or failure in the honey crop of the next year, so far as the activity of the colony itself is concerned. Most of these condi- tions are under the control of the beekeeper, and consequently may be provided by him at the proper time. In this same period the bee- keeper must also guard against swarming. During the major period, moreover, surplus honey may be consumed needlessly by the colonies instead of being stored, if the colonies did not reach their maximum field strength for the season until after the honey flow was over. During the initial expansion there is a tendency for all colonies to increase brood-rearing activity regardless of conditions; during the final phase there is an irresistible tendency for all colonies to contract THE BROOD-REARING CYCLE OF THE HONEYBEE 11 brood rearing, but only during the major period do colonies exhibit such diverse characteristics in brood rearing as to indicate a response during this phase to conditions which serve to counteract a normal seasonal tendency. It therefore follows that during the major period the beekeeper by his manipulations can best modify the behavior of a colony in the direction which he most desires. PERIOD OF FINAL CONTRACTION The period of final contraction represents a continuous decrease in brood rearing during the end of the active season, until by the begin- ning of the inactive season brood rearing has ceased entirely. A con- traction of brood rearing is a normal occurrence before the winter sus- pension, and is purely a seasonal phenomenon. The decrease may be abrupt, dropping from a high rate of activity to zero in a remarkably short time, as happens in regions characterized by short, active seasons with overlapping honey flows. In regions where there is scarcely any incoming nectar during the latter part of the active season, the final contraction may not be pronounced in its last stages. In short, the rapidity of this decrease is dependent upon the proximity of the last honey flow to the period marked normally by a complete suspen- sion of brood-rearing activities. The greater the quantity of sealed brood in the hive when the seasonal contraction begins and the nearer in time this beginning is to normal seasonal suspension, the greater are the chances for successfully passing through the inactive season, because such a condition in any colony means that it will enter the ■winter period with far more young bees than will one in which the final seasonal contraction is gradual and covers a relatively long time. SEASONAL SUSPENSION In the period of seasonal suspension a complete cessation of all brood-rearing activities takes place in a normal colony which is wintering well. Any brood rearing which occurs during this period is out of season, being abnormal and the result of some harmful factor, such as poor stores, an insufficient number of bees, insufficient insula- tion, or some outside disturbance of the colony itself. The length of the period of seasonal suspension varies greatly, according to the length of the winter. To bring his colonies through this period successfully is one of the most important problems of the beekeeper, in warm as well as in cold climates. As a matter of fact the problem is often more acute in regions with short inactive seasons than it is elsewhere, not only because there are more flight days but also because there is a less abrupt break from a high level of brood-rearing activity at the end of the previous season, so that colonies under such condi- tions will have fewer bees at thie beginning of the inactive season. Although the large number of flight days is an advantage in connec- tion with the more frequent possession of a poorer grade of honey stores in such localities, it is a disadvantage in view of the fact that useless flights throughout the inactive season rapidly deplcto the population of a colony which entered the period of suspension of brood rearing under less favorable circumstances than normally is the case in regions with short active seasons. 12 BULLETIN 1349, U. S. DEPARTMENT OF' AGRICULTURE DESCRIPTION OF THE COLONIES USED IN 1921 The brood-rearing investigation in 1921, as originally planned, was to be carried out along lines which would tend to show the effects of insulation, of stores, and of the age of the queen. Accordingly, each of the 16 colonies used in 1921 had been wintered in two 10-frame Langstroth hive bodies, eight colonies having been left all winter without packing and eight having been wintered in quadruple packing cases. The colonies were not in any way manipulated for the purpose of changing their brood-rearing rate, except that the addition of frames or supers had some influence in this respect. The record is, therefore, largely a presentation of what bees do without the interference of the beekeeper. The colonies without packing comprised two groups of four, and it was the original in- tention that the packed colonies should comprise the colonies in two packing cases. Since, however, one of the packed colonies became queenless during the winter, a colony packed in another group was used in its place. The group containing three colonies was unpacked on March 8, while the substituted colony (No. 10) was unpacked on March 21; but the group containing four colonies was not unpacked until May 5. Three of the colonies without packing had queens bred in 1919, whereas all the others, including both those with and without packing, had 1920 queens. Lack of stores was not during 1921 a factor in the brood-rearing activity of the eight unprotected colonies, as each colony proved to have more than sufficient stores for all purposes. Of the colonies packed all winter, four had heavy stores of honey, whereas the other four had light stores, all in the second hive body. The early spring was so favorable for incoming nectar and pollen, however, that each colony except No. 15 had sufficient stores for spring brood-rearing purposes. All the 16 colonies used were located at the Bee Culture Laboratory, Bureau of Entomology, at Somerset, Md., near Washington, D. C. In an endeavor to determine some of the factors which influence brood-rearing activity during the three phases of the active season, it is of interest to study the seasonal brood records of these 16 indi- vidual coloniek. For this purpose the general features of the brood- rearing activity in 1921 of 15 colonies are presented here, as well as a somewhat more detailed study of the brood record of an additional colony for two successive seasons. For each of these colonies a seasonal curve (figs. 1 to 16) has been constructed, based on counts (Tables 1 to 16) of sealed brood made from photographic records taken in 1921, as well as an additional curve (fig. 17) for one of these colonies, based on data (Table 17) from 1922. Although at first glance the 17 curves seem to show little correlation, they reveal definite relationships on closer study. The apparent differences result from abnormal conditions within individual colonies, which caused modifications in the brood-rearing responses. On the whole, the 17 curves point to certain more or less definite and constant seasonal variations in brood-rearing activity, due to normal, seasonal stimuli, but subject to modifications by the presence of any abnormal factors. This fact becomes apparent upon an examination of the two successive seasonal curves presented for colony No. 4. It may be noted in passing that, even before unpacking colonies wintered in two hive bodies in quadruple packing cases, removal of frames from the lower hive bodies can be accomplished readily by THE BROOD-REARING CYCLE OF THE HONEYBEE 13 the following method: Before packing, three frames in the lower hire body are replaced by two chaff division boards the lugs of which have been sawed off. After removing the frames in the second hive body these division boards in the lower hive body can be pulled out easily, and the space created by their removal is then sufficient to allow removal of the frames which were there with them. The space occupied by two packed division boards of the type commonly manufactured is equivalent to that occupied by three Langstroth frames. SEASONAL CHARACTERISTICS OF 1921 In the season of 1921 the weather conditions at the location of the bureau apiary were not favorable for a maximum honey crop. Although warm weather in late February and the fore part of March had brought out fruit bloom and other flowers somewhat prematurely, and although the thermometer registered as high as 90° F. (32.22° C.) on March 27 and 28, on March 29 and 30 most of this early bloom was destroyed by frost. On the morning of April 1, moreover, traces of snow were visible on the covers of the hives. During the fortnight beginning March 29, with one exception,- when 59° F. (15° C.) was the minimum registered, the temperature dropped each night well below the clustering point (57° F., 13.89° C.) even reach- ing the freezing point on six occasions. One 96-hour period had a maximum of only 62° F. (16.67° C.) and a minimum of 29° F. ( — 1.67° C.). On four occasions 57° F. (13.89° C.) was the highest tempera- ture recorded during a 24-hour period. Slight precipitation occurred on eight days of this fortnight. Such weather curtailed pollen and nectar gathering, which had just before been going on very actively. Warmer weather set in again with the middle of April, and con- tinued until the end of the summer, although there were a few days of cold, rainy weather in early May. The temperature during April ranged from 29° F. (— 1.67° C.) to 96° F. (35.56° C.) ; in May, from 39° F. (3.89° C.) to 93° F. (33.89° C.). Rain on six consecutive days, beginning with May 11, spoiled the chances for a large yield from black locust {Robinia pseudacacia) . On May 18 the tuliptree ( Liriodendron tulipifera) began to yield nectar, but a four days’ rain beginning May 23 put an end to nectar from this source. Nectar was available in small quantities frorn^ other sources during the latter half of April and throughout May. During practically the whole of June much honeydew, as well as pollen, was available, and for a short time after June 15 a slight amount of nectar from basswood ( Tiiia spp.) and sweet clover ( Melilotus alba) was collected. During July little nectar came in, although during the week beginning July 15 a small quantity of pollen was brought into the hive. Beginning August 3, and through- out the rest of the month, pollen was carried into the hives in largo quantities, and it was fairly abundant during September. Beginning September 12 and continuing until the end of that month a small nectar flow from various Compositae was on. October proved to be a period of little activity as far as pollen or nectar gathering was concerned. 14 BULLETIN 1349, U. S. DEPARTMENT OP AGRICULTURE BROOD REARING OF A TYPICAL COLONY FOR TWO SUCCESSIVE SEASONS Of the 16 colonies used in 1921 the seasonal brood-rearing record of one (No. 4) during 1921 and 1922 has been chosen as normal and as typical of the whole group. This colony stored the most surplus honey during the two successive seasons and its seasonal brood- rearing curves for the two years are strikingly similar, taking into consideration certain minor differences due primarily to weather con- ditions. During the winter preceding each of the two seasons this colony was left without packing in two 10-frame Langstroth hive bodies with abundant stores of honey. Nothing additional was done to stimulate brood rearing in the course of the two years except to furnish an abundance of room at all times, the queen having free access to all hive bodies. In 1921 the first super, or third hive body, was given to the colony on April 26, whereas in 1922 the first super was given on April 14. In both seasons the queen still had plenty of room in the original two hive bodies at the time the third hive body was given. Other supers were given later, so that in 1921 the colony had a maximum number of five hive bodies, and in 1922 a maximum of six. Owing to the policy adopted of providing an abundance of room at all times, not only for incoming nectar but for the egg-laying activity of the queen as well, the total number of hive bodies furnished during each season provided one hive body in excess of the actual minimum requirements of the colony. No added hive bodies were removed until October in either year; in 1921 the number was reduced to two on October 11, and in 1922 the same was done on October 16. Although there was no restriction to any possible expansion of the brood area, on no occasion was brood found above the lower three hive bodies. The same queen was used throughout both seasons, having been introduced into the colony in late summer in 1920 as soon as she had commenced to lay. For purposes of identification her left wing was clipped in July, 1921. In brief, each spring found this colony with a fairly strong force of bees, a prolific queen, combs composed chiefly of worker cells, and no shortage of honey stores. INITIAL EXPANSION, COLONY NO. 4 In each of the two years of this experiment, as shown by obser- vations, the chief source of nectar, tuliptree, was yielding freely by May 21. For these years, then, the maximum amount of sealed brood should have been attained during the last week in April, but in neither 1921 nor 1922 did the maximum brood-rearing correspond with ideal conditions. (Figs. 4 and 17 and Tables 4 and 17.) In 1922 the maximum amount of sealed brood was reached during the first week in May, whereas in 1921 this maximum was not attained until about two weeks later, although a more auspicious beginning had been made occasioned by the unusually early spring. Nectar and pollen came in abundantly throughout March, but the secretion of nectar and the production of pollen were affected adversely by the inclement weather at the end of March and beginning of April, and probably the lower temperatures also affected unfavorably the activity of the unpacked colony. The adverse conditions finally checked the initial seasonal expansion of the brood, as is shown by a decrease in the amount of sealed brood in the latter hah of April, THE BROOD-REARING CYCLE OF THE HONEYBEE 15 1921. A recovery in the rate of brood rearing was made subse- quently, but at a time when factors associated normally with the major period were making themselves felt. It follows that the maximum amount of brood rearing in 1921 was not purely the re- sult of the initial seasonal tendency. In 1922, on the other hand, there was a late spring, inclement weather in early March causing a temporary shortage of pollen in the hive. These conditions restricted somewhat the initial seasonal tend- ency, as is clearly evidenced in the brood curve for that month. In spite of this beginning, April weather proved so favorable that by the end of the month the principal sources of nectar were as far advanced as in the previous year, and brood-rearing activity became so pro- nounced that the maximum for the season was practically reached before factors peculiar to the major period became dominant. In fact, the maximum in sealed brood in 1922 was attained in advance of the tuliptree honey flow, and about two weeks in advance of that of the previous year. THE MAJOR PERIOD, COLONY NO. 4 In 1922, the more typical year, the beginning of the major phase was marked by a maintenance of brood rearing at the highest rate of the year (fig. 17). Since the maximum had been attained just before the locust bloom, the high rate was kept up for a couple of weeks. During the week of maximum sealed brood, brood rearing was undoubtedly still being carried on under the impulse of the intitial tendency, but influencing factors characteristic of the major period were also becoming evident. The week of the maximum marked the point of division between the initial seasonal tendency and the major period. Nectar subsequently coming in from the tuliptree tended to restrict the queen, and after this honey flow there was a dearth of nectar until the middle of September. During June, however, there was an appreciable amount of incoming pollen, and in August there was an intense pollen yield. As a consequence the decline in brood-rearing activity which set in with the beginning of the tuliptree honey flow and extended until the intense pollen yield in August was broken by a slight increase in June, in response to the pollen yield. The increase in August, on the other hand, was very pronounced. This was followed by a decrease until incoming nectar, chiefly from goldenrod, made itself felt by another slight in- crease in the brood area. In the main, brood-rearing activity during the major period of 1921 was very similar to that for 1922. In the month of June, 1921, the increase m brood rearing was more pronounced, for two reasons. In the first place, instead of a gradual decline following the peak, as in 1922, a sharp decrease occurred in 1921. This happened because just prior to the honey flow the queen had ascended to the third hive body to lay, there being already at that time an extensive brood area in this hive body. Incoming nectar, however, so quickly cut down the number of cells available for the queen as to force her soon to return to the second hive body. Here, too, so many available cells within the brood area proper had been filled with nectar during her absence that the total number of cells made empty either by emerging bees or by consumption of stores did not suffice during that week to permit keeping up her former rate. During the next two 16 BULLETIN 1349, U. S. DEPARTMENT' OF AGRICULTURE weeks, with the emergence of many bees in both the second and third hive bodies, the queen had more room and was able to approach a rate comparable to that attained in the same period of the following year. In the second place, there was an exceptionally large quantity of honeydew available, associated with incoming pollen during June, and also a certain amount of nectar from sweet clover. Throughout this month the queen had all the room needed for a normal response to these stimuli. In both seasons, after July 1, the brood curves follow parallel courses during the remainder of the major periods. At the end of July in each season brood-rearing activity had been reduced to approximately one-half that represented by the maximum of the same year. FINAL CONTRACTION, COLONY NO. 4 In each year the final seasonal contraction in brood-rearing activity took place almost entirely in October, covering only three weeks in 1921 and four weeks in 1922. The abruptness of the contraction in these few weeks is shown from the fact that in the last week of September there were practically half as many cells of sealed brood as were found in the maximum counts for the respective years. As a result, the colony entered each following season strong in bees. The brood-rearing record of this colony, although not ideal, is the most satisfactory of any of the 16 colonies because the maximum brood rearing bears some correlation to the initial expansion. The portion of the major period immediately following the period of main nectar secretion is not marked by a disproportionate degree of brood-rearing activity. In the late stages of the major period, more- over, there is an increase in brood rearing, providing a sufficient number of newly emerged bees at the beginning of the final contrac- tion to insure successful wintering and an auspicious beginning of the next active season. That conditions within the colony remained nearly constant during the two consecutive years is indicated by the striking similarity in the curves of brood-rearing activity during both active seasons. (Fig. 18.) At the beginning of the experiment this colony was fairly strong; and, although it was subjected to no special manipulations except to have plenty of room and stores available at all times, it was fully as strong in bees at the beginning Of the seasonal suspension in the fall of 1922 as it was in early spring in 1921. The performance of this colony, therefore, leads to the con- clusion that, other conditions being equal, a strong colony tends to remain strong. GENERAL OBSERVATIONS ON THE REMAINING COLONIES Having given the record for colony No. 4 in some detail for the two seasons, it is not necessary to discuss so minutely the records of the remaining colonies observed during the season of 1921. Only the outstanding points regarding the various colonies will be con- sidered. Unless otherwise stated, all colonies lived through the winter of 1921-22. Colony No. 1 had been wintered without packing but was provided with an abundance of stores and had a 1920 queen. The brood- rearing activity of this colony furnishes a good example of the re- sponse of a mediocre queen to such a combination of factors as suf- ficient stores, sufficient worker bees, and sufficient brood cells at the THE BROOD-BEARING CYCLE OF THE HONEYBEE 17 beginning of the active season. During the initial expansion a rate of brood rearing was reached almost equivalent to that at the height of the whole season. In fact, the amount of sealed brood (fig. 1 and Table 1) indicates no striking fluctuations in the rate of brood rearing during the three months when this was most active. Even the continued addition of many young bees to the population of the colony during this time failed to heighten the brood-rearing activity. The cold weather of April, the nectar flow of May, the honeydew and pollen yield of June, the pollen of August, and the nectar flow of September, may all be traced by the variations in the quantity of sealed brood. The response to the stimulus in August was only slight. In relation to the number of bees reared during the major period, the initial expansion was not nearly enough pronounced, nor did the final contraction represent a large enough break in brood- rearing activity to insure successful wintering. This colony died during the winter of 1921-22, leaving only a small quantity of stores in the hive. Colony No. 2 also had a 1920 queen, plenty of stores, and had been wintered without packing. Although this queen was more prolific than the queen in colony No. 1, as is shown clearly b}^ the curves of sealed brood for these colonies (figs. 1 and 2, Tables 1 and 2), there were not sufficient bees in the colony at the beginning of the active season to cause the maximum brood rearing to be correlated closely with the initial expansion. The maximum was reached only in June, after the population of the colony had increased sufficiently over that obtaining during the initial expansion to take care of an enlarged brood area. This rate was reached coincidently with the honeydew yield in June. As the incoming honeydew was placed within the brood area, restricting the number of cells available to the queen, an abrupt decrease in brood rearing followed in the week immediately after the maximum. A partial recovery in the rate of brood rearing occurred along with a relatively small pollen yield in July, after which, with one exception, brood-rearing activity decreased continuously until the September nectar flow. This exception took place during the pollen yield of August, as is shown by the fact that at this time the curve remained at about the same level for one week. The brood- rearing activity of this colony shows the same responses to weather and incoming pollen and nectar as have been noted for colonies 1 and 4, there being, of course, differences in degree. The maximum brood-rearing activity of the season came after the major period was well advanced, and is consequently too much out of proportion to the initial expansion to represent ideal conditions. Colony No. 3 was wintered without packing, had been given heavy stores, and had a 1919 queen. This queen had enough bees at the beginning of the active season to support her maximum egg-layin" activity, as is shown by the fact that scarcely any more sealed brood was found in the hive on any occasion in May than had been found during the initial expansion in March and. April. (Fig. 3 and Table 3.) Because the queen was old and about worn out, incoming nectar in May soon severely restricted her activity, the rosul t being a marked drop in the brood rearing for that month. At the beginning of May she had been crowded out of the second hive body, in which she had been laying almost exclusively, into the lower hive body. Even there 46969°— 25f— Bull. 1349 3 18 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE the brood area was quickly hedged about by incoming nectar, with the consequent great reduction in cells available for egg laying. Emerg- ing bees made more room available in June; and consequently, under the stimuli of incoming pollen and honeydew, the queen’s activities were increased to a limited extent. After this there was a continuous decrease in brood rearing until the old queen was superseded in August. The colony had been endeavoring to super- sede her since May, but all queen cells had been removed as soon as found. Finally one was left intentionally, from which a virgin emerged. The colony had become rather weak by August, how- ever, and the curve shows no apparent response to the pollen yield of that month, although the old queen was still alive. The new queen had mated and had commenced to lay by August 25. During the nectar flow of September she attained a rate which compared favorably with that attained by the old queen even during the initial expansion. Owing to the fact that the old queen was failing, the brood-rearing responses of this colony to nectar and pollen yields are not so clearly shown as in colonies with more vigorous queens. The effect of the April weather is, however, very apparent. Colony No. 5 (fig. 5 and Table 5) was wintered without packing, had a 1920 queen, and was provided with abundant stores of honey. In fact, the second hive body was so well filled with honey that the queen was cramped for room in late March, and it was deemed advisable to replace two frames of honey in the second hive body with empty frames. Although before these empty frames were substituted the queen had reached the limit of room for egg laying in the second hive body, scarcely any brood was reared in the lower hive body until May Or June, even though plenty of room was available there in March. Before the cold spell in April there were more than sufficient bees in the hive to take care of brood in all available cells in the second hive body, but not enough to maintain in addition a large brood area in the lower hive body. The addition of the two extra frames permitted an expansion of the brood area in the second hive body, and the amount of sealed brood mounted slowly, even during the cold weather. During this period, however, the queen laid no eggs in the lower hive body. In her activities the queen kept pace with the ever-increasing num- ber of young bees from the first part of April until the maximum of the season. During this period, nevertheless, brood-rearing activity suffered a slight check in May, at a time when much pollen and some nectar were coming in. No super had as yet been given the colony, and, as much of the pollen and nectar were being deposited within the brood nest proper, the queen was once more cramped for room. A super given at this time relieved the shortage of room and fur- nished storage space for what little surplus was gathered during the tuliptree nectar flow immediately following, and more particularly during the honeydew yield of June, when the maximum brood- rearing activity of the season was attained. During the dearth of nectar immediately afterwards, brood-rearing activity fell off notice- ably until stimulated by the pollen yield in August. From the height of the response to this stimulation until the beginning of the Sep- tember nectar flow, brood-rearing activity underwent another strikmg decrease. There was somewhat of an increase in September, which, although rather conspicuous on the curve, was not nearly great THE BROOD-REARING CYCLE OF THE HONEYBEE 19 enough to insure the best wintering conditions. The greatly dimin- ished brood-rearing activity during the latter part of the major period must be attributed to a failing queen, because at any time after July, except for that drawback, there were sufficient bees, stores, and room to have resulted in a much greater amount of brood during August and September. Seasonal brood-rearing activity of the type represented by this colony is anything but satisfactory from the standpoint of obtaining a honey surplus. Owing to lack of room, the initial expansion did not have a chance to proceed normally, even though sufficient bees to meet the queen’s egg-laying capacity were not on hand. As a result of having its early development arrested and retarded, the initial expansion became merged with the major period. Owing to lack of sufficient bees, brood-rearing activity increased with relative slowness, even after sufficient room had been provided. As a conse- quence, the maximum of sealed brood was not attained until too late for the resulting bees to be useful in gathering nectar for surplus. Comparatively slight brood-rearing activity at and preceding the final contraction spelled danger to the colony in the coming winter, and well illustrates the evil results of failure to requeen at the proper time. This colony died in the winter of 1921-22, with some honey still remaining in the hive. Colony No. 6 (fig. 6 and Table 6) had a 1920 queen, stores in abundance, and had been wintered without packing. For this colony, there is no sharp distinction between initial expansion and major period. Colony No. 6 had an overabundance of stores in the second hive body in early spring, thus restricting the queen; but it did not have enough bees to permit expansion of the brood area downward into the lower hive body at the rate at which it was begun in the second hive body. There occurred, therefore, a slight diminu- tion of the brood area at the end of March, but the colony so increased in population during the fore part of April as to provide more than enough bees to take care of brood in all the cells available. Brood- rearing activity consequently increased to the maximum in late May, excepting another slight break in April, caused by weather conditions. Following the maximum there was a rather abrupt decrease due to the presence of fresh nectar temporarily within the brood nest. A super added during the last week of May partially provided room for the honeydew in June, and thus tended to eliminate further restriction of the brood area. In fact, an increase in brood rearing took place in June. Although the queen had reached in May her maximum for the year, she was able to make a noticeable increase in her rate, even in July, pfter having dropped off from her maximum for June. It may be added that some pollen was coming in at this time. A decrease then followed which lasted until September. Possibly owing to the July increase, the response to the pollen yield of August is apparent only as a slight diminution in the rate of decrease. IIow- ever, a pronounced increase in brood rearing took place during the September nectar flow. The initial expansion of the colony is merged too completely with the major period. During the major period there was too much brood rearing, resulting in a useless con- sumption of stores. Colony No. 7 (fig. 7 and Table 7) had a 1919 queen and sufficient stores and had been wintered without packing. Sufficient bees were 20 BULLETIN 1349? U. S. DEPARTMENT OF AGRICULTURE in the colony at the beginning of the initial expansion to have sup- ported a greater rate of brood rearing than was attained. Sufficient cells were also available, but the queen was evidently too old to have made any better showing than she actually did. Rer maximum was reached during the initial expansion. A rate nearly equal to the maximum was maintained for about a month, and then a decline set in which reduced brood-rearing activity to practically zero by the end of August. At that time the old queen was superseded. The brood-rearing activity of the new queen, even in September, equaled that of the old queen during the initial expansion. The fact that tills colony at the beginning of the active season did not have a queen prolific enough to allow it to carry on brood rearing at a rate con- sistent with its strength in bees, available brood cells, and honey stores, accounts for the fact that it does not exhibit all of the responses to seasonal phenomena found in the other colonies. Colony No. 8 (fig. 8 and Table 8) also had a 1919 queen, had been without packing all winter, and had plenty of stores. This colony had the poorest queen of any of the 16 colonies used. During the initial expansion she attained almost her maximum rate for the season. The cold weather in April caused a slight decrease, but her maximum was reached in early May. Incoming nectar in that month restricted her activity and a decline followed. On August 18 the colony was queenless. A virgin queen emerged during the next week but never mated. Finally on September 8 a young queen was introduced which began to lay on September 15, but was lost two weeks later, after having made a good start. Another queen was introduced successfully in October, but too late to produce much brood. Colony No. 9 was unpacked on March 8, nad sufficient stores and a 1920 queen. The curve of sealed brood for this colony (fig. 9 and Table 9) is typical of a queen which lays at her maximum rate during the season, the rate being fairly uniform during most of the major period. As this was a packed colony, only seven frames were in the lower hive body. The three frames completing the normal number were not added until the last week in March, and the colony became somewhat crowded for room, thus restricting the queen during the period of initial expansion. The added combs helped to relieve the brood area proper from further restriction by pollen, with possibly a little nectar. The fact that the queen was utilizing only the second hive body at the time of the inclement weather in April, coupled with the fact that there were more than sufficient bees on hand to allow an expansion of the brood area even at this time, resulted in an increase of brood rearing during April until the maximum of the season was reached at the end of the month. From that time until late August, when this queen was naturally superseded, brood was reared at a fairly uniform but generally decreasing rate. Although an excess of room was provided, so that this queen was restricted in no respect, there were but slight reactions to the nectar of May and the pollen and honeydew of June. Supersedure interfered with the response to the pollen yield of August, although there are indications that the re- sponse had already begun before the new queen emerged. She was laying by the beginning of September. The increase in brood rearing during that month was due probably not only to the incoming nectar but to the presence of a young queen as well. Criticism of the THE BROOD-REARING CYCLE OE THE HONEYBEE 21 seasonal brood-rearing activity of this colony may be put in two ways. It may be said either that brood rearing during the initial expansion did not sufficiently exceed that maintained during the larger part of the major period, or that brood-rearing activity during the larger part of the major period was continued too nearly at the same rate as during the initial expansion, thus causing too large a consumption of stores in the hive in rearing brood uselesssly and in feeding an idle population. Colony No. 10 (fig. 10 and Table 10) began the season under ideal conditions as far as the queen herself was concerned, and was un- packed March 21. The original queen had been introduced in 1920, had plenty of bees, room, and stores, and consequently attained her maximum rate of brood rearing during the period of initial expansion. The weather apparently had little effect on the activity of this queen; the decline from the maximum of the season is rather gradual and the subsequent increase is made slowly. The colony had plenty of room, but the nectar in May received a relatively slight response. The queen was not seriously restricted by incoming nectar, and probably was laying at her maximum capacity. A further slight response was brought about by honevdew in June. In July, however, the queen was lost and brood rearing dropped off abruptly. A virgin queen was reared which mated and began laying in August. This queen soon reached the limit of empty brood cells, as the brood nest had become rather well filled with pollen and some nectar. Two empty frames were therefore placed in the second hive body, which were promptly used by the queen. As a result of this additional room the queen had all the cells necessary during the remainder of the season. The initial expansion of the season was timely enough, but it was not suf- ficiently greater in activity than was brood rearing during June and July to spell success in surplus. The final contraction presents fairly satisfactory conditions. Colony No. 11 (fig. 11 and Table 11) afforded the best illustration of any of the 16 colonies of what may be accomplished during the initial seasonal expansion if conditions within the hive are favorable. This colony had a 1920 queen, was unpacked on March 8, and had no more stores than were sufficient to meet brood-rearing require- ments before early nectar began coming in; a condition which gave the queen a maximum amount of room. Besides these factors, there were at the beginning of the active season plenty of bees and a pro- lific queen. Under the spur of the tendency toward the initial expansion, the brood area increased with great rapidity. At the time of the change in weather conditions in April, the brood area had been expanded so far that a further expansion was not possible and a break occurred. A recovery was soon made which culminated in May, when incoming nectar caused another restriction of the brood area. During June a short-lived recovery was made which was terminated by a decrease, probably due in part to a restriction of the brood area by honeydew, but more likely to the absenco of any great stimulus toward increased brood-rearing activities during July. Even so, the queen continued active enough to respond in a small degree to the minor pollen yield in July. Owing to the large number of field bees on hand in August, tne brood area was so restricted by the quantities of pollen brought in that the queen was unable to make much of a response. In fact, a rapid decrease followed 22 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE this pollen yield, and continued until the fall nectar flow. Increased brood-rearing activity under this new stimulus was slight in com- parison with that under stimuli earlier in the season. Notwith- standing the fine example of an optimum initial expansion repre- sented by this colony, brood rearing was a little too active in pro- portion in May, and especially so in June, and exhausted uselessly what little nectar had been gathered in May. Colony No. 12 had a queen which attained her maximum capacity during the initial expansion. This colony was unpacked March 8, had a 1920 queen and stores just sufficient, leaving plenty of room available for brood rearing. The April decline, characteristic of so many of the other colonies, appears to have been for the most part avoided in this colony, probably because of the presence of a large number of bees. Although April weather may have prematurely ended the initial expansion, the fact that, excepting one sporadic occasion, the quantity of sealed brood did not at any time equal the quantity found at the height of the initial expansion, tends to show that the queen was laying throughout at about her maximum capacity. The curve of sealed brood (fig. 12 and Table 12) shows a decrease in early May, followed by an increase later in the month which may be correlated with the tuliptree nectar flow. The in- crease, however, did not bring brood rearing up to the maximum attained during the period of initial expansion. A decrease followed the May increase; but it was checked in part by the incoming honey- dew and pollen, after which there was a sporadic upshoot at the very end of this period. From that time on there was a marked, rapid decrease in "brood rearing until the queen was superseded in August. The new queen was laying on August 16 and quickly attained a relatively high rank in September, but unfortunately was lost in that month. A new queen emerged and mated but did not begin to lay until October 4, too late to make much of a showing. The two peaks in the curve, following the maximum of the initial expansion, disturb the proportions of an ideal curve; but the bees reared during these two increases in the brood-rearing rate doubtless enabled the first new queen to establish the good record to her credit. Colony No. 13 (fig. 13 and Table 13) was unpacked on May 5 and had a 1920 queen and more than sufficient stores. At the time of the bad weather in April practically all available cells were in use either for stores or for brood, the lack of available brood cells accounting largely for the April decline shown. A super was added on April 29 which, with three frames placed in the lower hive body instead of the packed division boards, provided plenty of room. Brood rearing then so increased that the maximum of sealed brood was reached the second week in June. In May incoming nectar interfered slightly with brood rearing just prior to the maximum, and honey- dew caused a further restriction just after the maximum. The rate then became stable for a few weeks until the queen disappeared. A virgin was reared naturally which mated and began to lay on August 19. Brood-rearing activity was carried on at a fairly rapid rate, but suffered a slight check just prior to the September nectar flow, incoming pollen causing a restriction in room. The peak attained in September was sufficiently high to insure plenty of bees for winter. As the maximum brood-rearing activity came in the major period after the main honey flow was over, there was no occasion to expect a large honey crop in 1921. THE BROOD-REARING CYCLE OF THE HONEYBEE 23 Colony No. 14 (fig. 14 and Table 14) bad neither sufficient bees nor sufficient room to allow the queen to lay eggs at her maximum rate during the initial expansion. This colony had a 1920 queen and an abundance of stores. The large quantity of stores, both in the upper and the lower hive bodies, reduced the area available for brood rearing to less than the requirement of the colony. This condition was rendered more acute by the fact that until the time of unpacking, May 5, there were only seven frames in the lower hive body. The effect of lack of room is visible in the brood curve at the end of March. At this time, to give more room, a frame in the second hive body was replaced by an empty brood frame, with the result that there was a slight increase in brood rearing. This extra space was not sufficient, However, for the incoming pollen and nectar and for the brood. A decline in April resulted from this factor and from any influence of the weather. A super was added in the last week in April, shortly before the colony was unpacked. This relieved the congestion, and in May the brood rearing rose well above the height attained during the initial expansion. That the maximum was not attained until June is evidence of the fact that there were not sufficient bees in the hive from the latter part of April, when an abundance of room became available. After June came the normal seasonal decrease in brood rearing, lasting until pollen in August produced a rather long-continued response. Following this another decrease occurred until goldenrod nectar made itself felt. Owing to a large number of field bees the brood nest became much hemmed in. The brood-rearing record of this colony was entirely unsatisfactory from the standpoint of the honey producer, because the height of brood rearing came after all chance of storing surplus was past, and was so pronounced that to maintain the colony’s new population meant serious inroads on what nectar had been gathered. Colony No. 15 (fig. 15 and Table 15) was unpacked on May 5, had a 1920 queen, light stores, and not sufficient bees to allow brood- rearing activity to keep pace with the egg-laying activity of the queen. This was the only one of the four packed colonies (Nos. 11, 12, 15 and 16) provided with light stores which showed any diminution in brood rearing directly traceable to lack of honey stores. When observations on this colony were first made in March, 1921, the honey stores were found to be low. Not being so strong in bees as the other colonies just mentioned, it was not able to gather so much ne ctar during March, and therefore had to use proportionately more reserve stores in that month. ' The records of sealed brood reveal a check to the brood-rearing rate in late March, followed by an increase. The fact that this increase took place within two weeks after giving the colony a full comb of honey in the lower hive body is an evident indication that, at the time the extra comb of honey was given, there was neither sufficient honey in the hive nor sufficient nectar coming in to support any great increase in brood-rearing activity. The increase was only short lived, because at tho time oi the cold spell in April the brood area had become as largo as could be covered by the bees in the hive during such weather. The queen was necessarily restricted to this area lor tho time being, and as a largo part of the brood in it remained sealed for several days she did not have sufficient cells available for further egg laying. The emergence of young bees, however, gave her a chance to refill the cells in this 24 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE area, as is shown by a rise in the curve during the first week of May. In the meantime, however, the upper edges of this area had been cut off by incoming nectar and pollen. Field bees had been kept busy during the latter part of April after the cold weather, although nectar was not very abundant. To relieve the resulting congestion, a super had been given in the last week of April, about a week before unpacking, and on unpacking three more frames had been placed in the lower hive body to take the place of the packed division boards. With this extra room the queen was able to expand the brood area in proportion to the number of nurse bees. The result was a sharp rise in May to a maximum which was maintained through- out most of June. During this period there was plenty of room in the hive to provide cells for the activities both of the field bees and of the queen. From the latter part of June until the pollen yield of August the activities of the queen became more and more restricted, excepting a brief response to pollen in July. A fair response was made to the pollen yield of August. By this time the large number of bees which had emerged during June, plus the brood reared after that month, had depleted the honey stores to a point which caused a serious diminution of brood-rearing activity between the pollen yield of August and the nectar flow of September. In the spring the colony had not suffered from want of stores but had no great surplus; during the May nectar flow and the greater portion of the honeydew yield in June sufficient field bees had not as yet developed to add much to the surplus stores. Consequently during the summer practically all of the honey stores were consumed. To save the colony it became necessary to add three full combs of honey to the second hive body in the second week in September. This factor, coupled with the oncoming nectar flow, caused another increase in brood rearing. By the second week in October honey stores were practically depleted once more, and it was necessary to add more irames of honey. Under such circumstances brood rearing had not progressed actively enough just prior to the seasonal contraction to afford this colony an optimum number of bees for winter. The whole curve is highly unsatisfactory, because the initial expansion does not represent sufficient brood-rearing activity and because too wide a gap separates the maximum activity of the major period from the initial expansion. At the end of the season there is also too wide a gap in the continuity of the curve for the major period between the point of demarcation of the seasonal contraction and the next pre- ceding high point of the major period. Colony No. 16 had a 1920 queen, light stores, and was unpacked on May 5. Although the brood curve for this colony (fig. 16 and Table 16) shows a rather rapid and early initial expansion, there were not quite enough bees for the queen. During the initial expan- sion the queen was utilizing all of the 10 frames in the second hive body. She had reached the limit of the area of brood which could be cared for in March, however, and a slight decrease followed. With the emergence of young bees more room was available within the brood area, and the queen began to take advantage of this, but was restrained somewhat by the cold weather in April. When the warm weather came, before the queen could take possession of all the ceils made available by emerging bees, a large number of cells had already been used for incoming nectar and pollen, resulting in a striking drop THE BROOD-REARING CYCLE OE THE HONEYBEE 25 in the curve. As a matter of fact, at one time the queen was confined to one side of the brood area by a comb filled almost entirely with fresh nectar. More room was given by the addition of a super dur- ing the last week in April. Unpacking on May 5 meant the addition of three frames in the lower hive body to replace the packed division boards. As a result, both of the extra room and of the stimulation of the May nectar flow, brood-rearing activity expanded back to the level attained in March. Incoming nectar occasioned a slight de- crease during the latter part of May, but feeding of larvae on hand soon removed this restriction. The maximum of the season was attained in June, but soon gave way to a sharp decrease brought about by incoming honeydew. A partial recovery was made, which in turn was stopped by another decrease because the brood area had been so restricted by honeydew that few empty cells were to be found. More room was available within the brood area proper at the time of the small pollen yield in July, which resulted in an increase followed by a decrease. Loss of the queen at this time then caused a suspension of brood rearing. A virgin was reared, which mated and began laying before the September honey flow. As in the case of colony No. 15, there is too wide a gap between the initial expansion and the main activity of the major period, and also too much of a break in the continuity of brood rearing just prior to the final con- traction for this colony to have had the proper population during the various phases of the active season. GENERAL DISCUSSION OF THE RECORDS FOR 1921 The brood-rearing records of these colonies show the region about Washington to be one in which seasonal brood-rearing activity tends toward slackening during the major period. The main brood rearing of the season comes before the occurrence of this tendency; but, fol- lowing it, brood-rearing activity increases in normal colonies suffi- ciently to provide an adequate number of bees for winter. INFLUENCE OF POLLEN AND NECTAR Throughout the season direct responses were made to incoming pollen and nectar. The main nectar flows of the year come during the forepart of the active season, which is also normally the time of greatest brood-rearing activity. During July scarcely any fresh nectar is found, and brood-rearing activity is greatly diminished. In September there is a nectar flow, and associated with it is heightened brood-rearing activity. The correlation between a good pollen yield and brood rearing is well illustrated by the expansion of the brood areas during the pollen yield of August. That there will even be a response at times to a light pollen yield is shown by the colonies (Nos. 2, 6, 11, 15, and 16) whose brood-rearing activity did not decrease continuously throughout July. For the most part, owing to abnor- mal circumstances, conditions within these colonies did not become conducive to maximum brood-rearing activity until just before the period of nectar dearth; as a result, only a slight stimulus was needed to create a response in such colonies. That definite brood-rearing responses will be made from year to year to certain constant seasonal stimuli, other conditions remaining equal, is well brought out by the brood records of colony No. 4 for 1921 and 1922. 46969°— 25f— Bull. 1349 4 26 BULLETIN" 1349, U. S. DEPARTMENT OF AGRICULTURE CONDITIONS WITHIN THE HIVE The value of insulation in early spring is not demonstrated clearly in the case of these colonies because virtually summer weather pre- vailed in March, thus producing conditions within the colonies which tended to offset in a measure any evil effects of the cold weather in early April. Unfortunately, too, the colonies (Nos. 13, 14, 15, and 16) which were left packed until May suffered from lack of room just at the time of the unfavorable weather, so that no comparison may be made with them. Of the colonies left all winter without packing four (Nos. 5, 6, 7, and 8) present such abnormal conditions due both to lack of room and to failing queens as to offer little light on this subject. The other four colonies, which were not packed for winter, however, did not suffer from lack of room, and, although each dif- fered as to prolificness of queen and colony population, each shows a break in brood rearing associated with the cold weather of the fore part of April. Colony No. 4, nevertheless, had sufficient bees to overcome quickly the effect of this weather and to proceed to the maximum brood-rearing activity of the year. Although the brood- rearing activity of three of the colonies (Nos. 9, 10, and 12) unpacked in March appeared to be somewhat restricted during early April, the later performance of the queens in these colonies and the fact that no great decrease in brood-rearing activity immediately took place, indi- cate that sufficient bees were on hand to keep up the temperature of the brood area in its entire extent at that time, and even to have cared for a larger area had the queen been capable of increased egg laying. The brood area of the other colony (No. 11) which was un- packed in March shows a decided restriction following the cold weather. Although a comparison of the direct effect of adequate insulation on brood rearing can not be made from the brood records of these colonies in 1921, the fact remains that the colonies minus packing which were most normal did suffer a setback in brood rearing as a result of the cold weather. A strong colony without packing, like No. 4, shows, however, a certain amount of resistance to the effects of such weather. CONDITIONS OF THE COLONY The influence on brood rearing of the three important factors, prolificness of queen, colony population, and brood cells available, becomes so interwoven in certain colonies that it is hard to trace the separate influence of each. The important part played by colony population in determining when the maximum brood-rearing activity of the season will take place is well illustrated by colony No. 2, which reached the maximum of the brood-rearing season relatively late. In this colony lack of population was beginning to restrict the brood ■ area somewhat, even in late March. Population of colony accom- plishes little in itself if the queen is not prolific. The brood records of colonies Nos. 3, 7, and 8 all attest this fact. It so happened that each of these queens had been introduced two seasons previously. On the other hand, the queen of colony No. 4, in 1922, the second season after her introduction, made as good a record as in 1921. This shows that in certain instances, at least, the value of a queen can not be determined merely by her age. THE BROOD-BEARING CYCLE OF THE HONEYBEE 27 STORES As already stated, none of the 16 colonies suffered from lack of stores at the beginning of the period of initial expansion. ' No data were obtained, therefore, showing the effect of want of stores on brood rearing. It was evident, however, as soon as nectar became prematurely available from fruit bloom, that certain colonies had such an abundance of stores as to result in lack of room for any expansion of the brood area. The history of colony No. 14 brings out this fact, although in this instance colony population also influ- enced the result. The importance of having sufficient stores in early spring until incoming nectar supplies the current needs of the colony, and the need of room sufficient to offer no check to the initial expansion, should emphasize to users of the Langstroth or other hive bodies of equivalent size the value of wintering a colony in two hive bodies, and show that otherwise a full population for the honey flow is liable not to be attained. OBSERVATIONS IN 1920 It has been thought of interest to introduce at this point the work done by LI03M R. Watson in 1920, while he was connected with the Bee Culture Laboratory. As already stated, through direct counts of all eggs, larvae, and sealed brood, he was able to obtain in that year a total of five seasonal brood records. (Tables 18 to 22.) It must be borne in mind that the curves (figs. 19 to 23) based on these records represent a total of all eggs, larvae, and sealed brood, whereas the curves (figs. 1 to 18) based on photographic records made by the writer represent sealed brood only. Furthermore, the curves for 1920 are not drawn to the same scale as are the other curves in this paper. These differences in themselves would be sufficient to cause the curves to present apparent discrepancies in the time and degree of response even to the same stimuli. Unfortunately, too few data on each colony are available, other than the actual counts of brood, to warrant a close correlation between the brood records for 1920 and those already discussed. Nevertheless, the results for 1920 are in line with those presented for the succeeding years. Each of the five colonies (A, B, C, D, and E) had been wintered in two hive bodies in quadruple packing cases. Colonies C, D, and E were unpacked on March 17, wdiile colonies A and B were left packed until April 13. In August, 1919, a young queen had been introduced into each colony. All the colonies were well provided with stores. In 1920, as in 1921, there was a period of cool weather during early April. From April 4 to April 11, inclusive, the minimum and maxi- mum temperature corresponded roughly with the freezing point (32° F., 0° C.) and the clustering point (57° F., 13.89° C.) respectively. In fact, with two exceptions, the minimum temperature on each day of this period was below 32° F. (0° C.) whereas on only two occa- sions docs the temperature record for any day within this period show a maximum of over 57° F. (13.89° C.). The two exceptions to the minimum are April 4, with a minimum record of 39° F. (3.89° C.), and April 5, with a minimum record of 43° F. (6.11° C.). The two exceptions to the maximum occurred on April 9 and April 11, the former having a maximum of 61° F. (16.11° C.), the latter of 64° F. (17.78° C.). The data available as to honey flows show that the most nectar of the year wras gathered during the last half of May. 28 BULLETIN" 1349, U. S. DEPARTMENT OE AGRICULTURE In 1920, contrary to conditions in 1921 and 1922, the nectar flow from black locust occurred during tbe last week in May, thus coming after the ’ tulip-tree nectar flow, instead of before. June also fur- nished either nectar or honeydew, and a slight amount of nectar was available in September. No data are available as to pollen yields. In the case of each colony, brood rearing in the spring of 1920 was begun first in the second hive body. A check to brood rearing took place during the cold weather of April in both the packed and un- packed colonies. Inasmuch as the method of direct counting neces- sitated keeping the frames out of the hive for a considerable time, thus creating a disturbance to the colony lasting over several hours, much brood must have suffered undue exposure, a fact which in itself would account for any check at this time, even though other data are not available. In colony E this check is not so evident, owing to natural requeening during this period. In May, in each case, before the maximum was reached incoming nectar gave a check to brood rearing, while the brood nest was being maintained in the second hive body only, sufficient bees not being on hand to engage in brood-rearing activity in the lower hive body. Consumption of this fresh nectar soon made more room available in the second hive body, and the con- tinued emergence of bees so enlarged the colony population that possession was taken of the lower hive body for brood-rearing pur- poses. Only one super was given to any colony, this being provided on May 21. In each colony the maximum brood-rearing activity of the season was reached shortly afterwards. During the last week in August no records were taken, a fact which accounts for the gaps in the curves at that time. Colony A (fig. 19 and Table 18), after the initial check to brood expansion in April, proceeded at a fairly rapid rate to its maximum. This was not attained until after the close of the tuliptree nectar flow. A rapid decline then ensued, lasting through June, probably first brought on by incoming nectar, but later accentuated by the fact that the queen was confined to the lower hive body by a queen excluder from June 9 to June 30. She was given the freedom of both hive bodies during the first two weeks of July, and a slight increase in brood-rearing activity took place. The dearth of nectar and pollen during the remainder of July brought on a further decline. On August 11 the super was removed. A slight increase in the brood- rearing rate was made in September at the time of the usual fall nectar flow. The queen was lost during the first week of October. A virgin queen was then reared and mated successfully, but too late to produce much brood before winter. The maximum brood rearing in this colony took place about a month too late for ideal conditions, and brood rearing just before the period of final contraction wTas not sufficiently great to afford an ideal number of young bees for winter. Colony B (fig. 20 and Table 19), after the checks of April and May, attained its maximum in early June, following which there was a slight decline due to incoming nectar. A recovery made in the latter part of the month was followed by the midseasonal decline. On July 9 the queen was confined to the lower hive body, and on August 12 the super was removed. A somewhat pronounced response was made to the fall nectar flow. Colony C (fig. 21 and Table 20) did less in brood rearing than any of the other colonies. The maximum rate was reached late in May. The midseasonal decline is not so sharp as in the case of the other THE BROOD-REARING CYCLE OF THE HONEYBEE 29 colonies ; in fact, a rate nearly equal to the maximum was maintained throughout June. On July 19 the queen was confined to the lower hive body. Little response was made to the fall nectar flow. The brood-rearing activity of this colony did not augur well either for surplus or strength for winter. Colony D (fig. 22 and Table 21) is the only other of the 1920 colo- nies comparable to colony A. It underwent the checks to brood rearing in April and May before reaching its maximum at the begin- ning of June. Colony A had reared more brood during this time than had colony D. Brood rearing in colony D, as in colony A, suffered a decrease immediately after the maximum. On July 7 the queen was confined to the lower hive body, was allowed the freedom of the second hive body on July 13, but was again confined to the lower hive body on July 20. Throughout the remainder of the season the brood nest was in the lower hive body. On August 10 the super was removed. During the last week of August the old queen was superseded. The combination of a new queen and the fall nectar flow caused a rather large increase in brood-rearing activity during September. Colony E (fig. 23 and Table 22) lost its queen in early spring and a laying queen was introduced, the result being a condition somewhat equivalent to an early spring supersedure. The giving of a laying queen undoubtedly made the break in continuity of brood rearing shorter than would have resulted had it been necessary to wait for the mating of a virgin queen so early in the year. Even so, the colony did not reach its maximum brood-rearing activity until the second week in June. A decline ensued, the sharpness of which indicates a restriction due to incoming nectar. A rather slight recovery was made during the last week of June, followed in turn by a sharp decline. On July 2 the queen was confined to the lower hive body; was given access to the second hive body again on July 10, but on July 16 was confined to the lower hive body once more, where she remained for the rest of the season. On August 13 the super was removed. This colony made little response to the fall nectar flow. In none of the five colonies did the initial expansion proceed without a check. The maximum brood-rearing activity of the season was reached too late to be of greatest value during the main honey flow, which, in the vicinity of Washington, D. C., usually occurs in May. With the possible exception of colonies B and C, in none of the colo- nies did brood rearing become so active just prior to the final con- traction as to insure the number of young bees needed for good wintering conditions. As in the curves for 1921 (figs. 1 to 16), so in the curves for 1920 (figs. 19 to 23) there is shown a tendency for brood-rearing activity in the vicinity of Washington to reach its maximum during the fore part of the active season; this is succeeded in turn by a midseasonal decline broken more or less by incoming pollen or nectar. This decline is usually checked somewhat during the latter part of the active season by an abundant pollen yield in August and a nectar flow in September. Thus, barring differences due to variation in seasons, strength of colonies, and certain other factors within the hives which would cause in different colonies a variation in the responses even to identical stimuli during the same year, the general character of brood-rearing activity -in the colonies under observation in 1920 is strikingly similar to that of brood- rearing activity in the colonies under observation in 1921 and 1922. 30 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE MIGRATIONS OF THE QUEEN WITHIN THE HIVE Besides the study of brood-rearing activity in the colony as a whole throughout the year, it is of interest to follow the brood-rearing activity of the same colony within particular hive bodies during that period, because, if adequate room is provided, one hive body is rarely the scene of the brood rearing of a normal colony throughout an entire season. The existence of such a piece of apiary apparatus as a queen excluder suggests how common an occurrence it is for the queen to transfer her egg-laying activity from one hive body to another. The causes of these vertical migrations and the ultimate effects on brood-rearing activity are as yet not fully determined. In passing, another type of migration should be noted, which takes place entirely within a hive body and which may be termed a hori- zontal migration, or a migration from frame to frame. A knowledge of the causes and effects of the queen’s migrations is of direct value in the determination of the size of frame and hive which will most directly contribute to a maximum brood-rearing activity. Since, all things considered, colony No. 4 was the most normal of the 16, the migrations of its queen in 1921 (Fig. 24) and 1922 (Fig. 25) will be considered somewhat at length. The cluster of this colony during the winter preceding each of the two active seasons was located in the second hive body. It is probably because of this fact that brood rearing began there each spring. During the initial expansion of each year, however, the queen approached the limit of cells available in the second hive body, whereupon the lower hive body afforded the only room for an enlargement of the brood area, because the first super had not then been added. In 1921, the season with inclement weather in April, there was comparatively little brood-rearing activity in the first hive body at any time. Such weather tends to contract the area occupied by the bees, and thus restrict the expansion of the brood area. These conditions, prevailing at the time of the queen’s first visit to the lower hive body, naturally caused her stay there to be rather brief. When the weather became better more room was already available in the second hive body, owing in part to a further consumption of stores. This additional space allowed the queen to increase her egg-laying activities without much enlargement of the brood area in the first hive body which resulted from her initial visit. Furthermore, almost immediately afterwards, in May, nectar began to come in rather abundantly and was deposited in the third hive body, which had now been put on, the presence of nectar in this super attracting the queen upward rather than downward. In the meantime the first hive body became well filled with pollen and nectar. It is interesting to note that in the latter part of April, in response to the inclement weather, the sealed brood curve for the second hive body (Fig. 24) remains near a certain level until the effects of the bad weather have ceased, and that in the first hive body this curve does not rise above the point marking its first appearance. In 1922 (Fig. 25) the inclement weather occurred before the colony was in need of expanding into the lower hive body. The queen w*as there- fore able to complete her stay below, and the sealed brood area in the first hive body came to occupy nearly as many cells at the end of April as had the brood area in the second hive body at the time of the expansion into the first. THE BROOD-REARING CYCLE OF THE HONEYBEE 31 The appearance of nectar in quantity in the third hive body is marked by migrations of the queen back and forth between the sec- ond and third hive bodies, her activity in the first dwindling away completely. In both seasons, by the beginning of the tulip-tree nectar flow the queen was at work in the third hive body, undoubt- edly having beeq drawn there by the presence of large numbers of bees and incoming nectar, as well as having found so many cells in the other hive bodies filled either with brood, nectar, or pollen. On each return to the second hive body she found deposited there much pollen and some nectar. Each season these migrations produced three clearly defined peaks of brood rearing in the third hive body. Owing to the fact that between these peaks nectar was being crowded as closely as possible around the brood area, even in the cells from which bees had emerged, the queen was restricted continuously in room for egg laying. As a result, each successive peak in the third hive body became smaller and in the end no eggs were to be found above the second. With the advent of the period of lessened brood-rearing activity in July, there was little incentive in either year for the queen to wander out of the second hive body. During the same period much of the honey and pollen in this hive body was consumed, with the result that at the beginning of increased brood-rearing activity during the latter part of the major period, caused by incoming pollen, there was sufficient room for a much larger expansion of the brood nest within this hive body than would be apt to take place at that time of year. This condition continued until the fall nectar flow, when the lateness of the season and possibly the slowness of the honey flow caused all incoming nectar to be deposited in the second hive body, around the outer edge of the brood nest or within it. The brood nest was then rapidly constricted by the moving of honey towards the center of the second hive body from both the first and third. Since in none of the other colonies were conditions such as to in- duce vertical migrations to such an extent as in colony No. 4, general conclusions are not in order at this time. When all space in the second hive body became filled with brood or stores, the queen in colony No. 4 went below because there was no other place to go, and the population of the colony was already sufficiently great to be using the lower hive body. As the season progressed and a super was added and occupied the queen went up into the third hive body also, where there were so many bees and so much fresh nectar. At no time did she show signs of deserting the second hive body; and, when forced out of the third hive body by nectar and kept out of the first hive body by pollen and nectar, she confined her activity to the second hive body for the remainder of the season. Although vertical migrations in brood-rearing activity were not carried to so extreme a degree by the other colonies as by colony No. 4, migrations did occur betweon the first and second hive bodies. In the case of the other colonies whose maximum brood-rearing activity came later in the season, consumption of stores had made more room available in the second hive body at the time of the need for the maximum room than was the case with colony No. 4, which needed room early. There was therefore not tho need for expanding the brood area into another hive body to such a degree as in colony No. 4. Without exception, however, each colony did maintain more 32 BULLETIN" 1349, U. S. DEPARTMENT OF AGRICULTURE or less of a brood area in tbe lower hive body during some portion of the season, and, in addition, colony No. 11 even made some use of the third hive body. In certain instances where much room was available in the second hive body horizontal migrations were found to exist. Whether vertical or horizontal, the peaks of brood rearing represented by these migrations are separated roughly by 3-week intervals, or the time necessary for brood to develop. Statements to the effect that the queen hesitates to cross from one hive body to another are often found in the literature of beekeeping. As far as colony No. 4 (figs. 24 and 25) is concerned, there is no indi- cation of any such hesitancy. It may be well to remember in this connection that Langstroth hive bodies were used in the experiment and that good worker combs were present. The curves at no point suggest that any of the breaks are due to a hesitancy in the trans- ference of egg-laying activity from one hive body to another; in fact, the sharpest rise in colony No. 4 during 1922 is accompanied by such a transfer with no resulting break. This holds true of other parts of the curve for the same colony in 1922 as well as 1921. It would seem, then, that the queen will readily ascend or descend from one hive body to another if the intenseness of brood-rearing activity necessitates more room, provided worker bees have taken possession of the other hive body either because of activities in connection with nectar gathering or because of an overflow population. COMPACTNESS OF BROOD NEST In spite of any migration of the queen within the hive, a study of the location of the sealed brood (figs. 26, 27, 28, and 29) throughout the active season shows a remarkable compactness in the brood area. At all times of the season, except at about the time of the final con- traction, the brood area of colony No. 4 occupied contiguous frames in the second hive body. The most apparent exception occurred during the final contraction in 1921, when the brood area in the second hive body became divided by combs which were filled entirely with honey. The same compactness is observable in the areas occupied in the first and third hive bodies, the only noteworthy exception being in 1922. In that year one of the frames in the third hive body became completely filled with nectar before the queen had occu- pied the frame on either side. WTen the brood area had included the frame next to this frame of honey, the queen passed around the full comb and laid eggs in the frame on the other side. For several weeks the brood area in the third hive body was thus divided. Wfienever the brood area crossed the limits of the second hive body into the first and third, this expansion took place almost entirely in territory as adjacent as possible to the second hive body. This is not apparent from the figures, because for each side of any frame the bar represents by its length the size of the sealed brood area in proportion to the total surface of each side of that frame, represented by the vertical dimension of the frame, and not the exact location of the sealed brood on that frame. Hence the sealed brood in each frame is represented dia grammatically as being in the center of that frame. In the figures only sealed brood is represented, while the present discussion refers to the compactness of all brood. THE BROOD-REARING- CYCLE OF THE HONEYBEE 33 The location of the sealed brood through both years brings out also the persistency of the brood area. By persistency of the brood area is meant a tendency to rear brood in the cells which have already been occupied by brood, and a tendency for any expansion in the brood area to take place only in cells on frames immediately adjacent to those already thus occupied. In each spritig the area of sealed brood was expanded rapidly, but during any expansion the area first used for brood rearing was kept occupied for that purpose. The rapid increase in the number of cells on each frame occupied by brood is just as striking as is the increase in the number of frames used. The first relinquishment of any part of the brood area for any purpose other than brood rearing was due to encroaching nectar. Through- out both years the second hive body maintained its predominance as the center of brood-rearing activity, even though at times the queen carried on extensive egg-laying activity in the other hive bodies. TIME RELATION OF BROOD REARING TO NECTAR GATHERED It was pointed out at the beginning of this bulletin that the honey crop may be reduced (1) by an insufficient number of worker bees, (2) by a consumption of surplus honey by bees reared out of season, or (3) by swarming induced by a congestion of bees in the brood nest. Because colony No. 4 stored more honey than any of the other 15, and because its brood-rearing activity during 1921 presented features more ideal for the region of Washington than did the others, only this colony will be discussed in detail from the standpoint of the time relation of brood rearing to nectar gathered. Although in 1921 the maximum of sealed brood in colony No. 4 came in conjunction with the height of the honey flow, the honej^ gathered during the main honey flow was not wastefully consumed by bees emerging later in the season, as the maximum field force was available to take advantage of an intense yield of honeydew. The maintenance for this large force during the summer was therefore provided through the efforts of the colony itself. In the season of 1922, although the actual maximum of sealed brood came a week late and a rate of emerging bees nearly equal to the maximum was maintained during part of the honey flow, the maximum rate had been nearly reached during the week prior to the peak. A creditable showing was therefore made, three supers being actually used for storing nectar, as stated earlier, whereas in 1921 one less super was given for this purpose. A swarming impulse was scarcely noticeable during either 1921 or 1922 in any of the colonies under observation. A few queen cells were started; following the prompt removal of these no further preparations for swarming were observed. In none of the colonies, however, was the queen allowed to be so restricted in egg laying at any one time as to result in any significant reduction in the actual number of young bees needed to care for larvae or perform other hive duties. This point is of importance because the presence of an overabundance of young bees has been held one cause of swarming. Dcmuth (7, p. 13) has stated: “ The fact that the tendency to swarm is greatest at about the time the bees arc rearing the greatest amount of brood has led to the belief that swarming is caused by the presence in the hive of a large proportion of young bees not yet old enough for field work.” But, since under normal conditions bees not of field age perform duties inside the hive, it would seem that the mere 34 BULLETIN 1349, U. S. DEPARTMENT OP AGRICULTURE presence of bees too young for field work does not in itself induce swarming unless there is such an excess of young bees beyond those required for hive duties as to interfere with the routine of the colony. Any appreciable excess of young bees arises, not for the sake of intensifying the natural swarming impulse, but rather as a result of other factors. Such an excess is bound to occur if conditions within the colony prevent a large number of young bees from performing the functions of their normal life cycle. This would be the case whenever brood-rearing activity reached the limit of cells available at a period when the brood-rearing area would have been further enlarged if more cells had been at hand, or if under similar conditions brood rearing were restricted through a reduction in the number of cells available for brood by their use for incoming nectar or pollen. The egg-laying rate being then reduced to a noticeable extent for several days, there would eventually be fewer larvae to care for, fewer cells to clean out, fewer cells to be sealed over, and a diminution in all of the activities incident to a period of intense brood rearing. In consequence, since such duties are usually performed by young bees, there would be many of the latter out of a job, so to speak. Throughout the observations in this research it has been noted that all colonies, strong as well as weak, tend to crowd incoming- nectar not only around the border of the brood nest but even within the brood area itself whenever an empty cell is found. Such a ten- dency has long been recognized. Consequently, if during a honey flow many cells within the brood area proper become available for depositing incoming nectar through the emergence of large numbers of young bees, the queen may be restricted suddenly in her activity, as happened in 1921 to the queen in colony No. 4. Under these conditions there would result an excess of young bees. Whether this excess would be large enough to induce swarming would depend on the degree and duration of restriction of the queen’s activities. In the case of colony No. 4 and the others the restriction was never of long duration. It has often been observed that just prior to swarming the queen almost entirely ceases egg laying, so that all unsealed brood disappears. Such a condition is simply the natural result if any large number of brood cells have been used for another purpose. Not only may the number of cells available for egg laying become insufficient through diversion to use for incoming nectar and pollen, but even without any reduction in number by these causes there will be too few whenever brood rearing itself reaches the limit of cells available in the hive. During a period of normal seasonal increase in brood-rearing activity, idle field bees may at times cause a congestion within the hive which apparently inter- feres with brood rearing. Any of these conditions may arise independently of any honey flow, and would tend to explain the fact that swarming is not always correlated with a honey flow. In the colonies under discussion a swarming impulse was doubtless restrained by the fact that at no time was any colony crowded for room long enough to cause a serious break in the continuity of brood rearing at a time when the tendency toward brood-rearing activity was strong. Whenever any queen became restricted in any particular hive body during a period favorable to continued, heightened brood- rearing activity, she was able to migrate to a more suitable region within the hive. Through this ability to transfer her activity else- THE BROOD-BEARING CYCLE OF THE HONEYBEE 35 where the queen was able to maintain an egg-laying rate which was normal in its response to the varying stimuli of the season as modified by the condition of the colony itself. Thus in these colonies the possibility of the occurrence of such a large excess of idle young bees as would be conducive to swarming was reduced to a minimum, a fact which may account largely for the absence of any indication of a swarming impulse in either season. EGG LAYING From the counts of sealed brood (Tables 1 to 17), the maximum daily egg-laying average over a 12-day period has been calculated for each of these colonies, 12 days being the average time repre- sented by sealed worker brood. Any daily average derived from sealed brood is not to be interpreted as representing the actual daily egg-laying performance of the queen bees in question, since it has long been recognized that a queen bee lays more eggs than ever develop into adults. This excess of eggs is often very evident in spring, becoming less apparent during maximum brood-rearing activity, and again becoming evident in the fall. The constant seasonal correlation found in the weekly counts of sealed brood throughout the years covered by this investigation shows, however, that a reliable index to seasonal brood-rearing activity may be ob- tained by counts of sealed brood. Since the success of brood-rearing activity is to be gauged by the number of adult bees reared, it is evident that this is more closely determined from counts of sealed brood than from any other type of brood count. In neither season did the queen in colony No. 4 approach any such a daily egg-laying rate as that found by Yon Berlepsch ( 3 ) in his exper- iment. The same holds true of the other 15 colonies studied in 1921. The highest daily average during any 12-day period, as derived from the counts of sealed brood, was found to be 1,587, and represents the performance of the queen in colony No. 4 in 1922, her highest similar rate in 1921 being 1,488. Of the other colonies in 1921, colony No. 14 had a maximum daily rate of 1,513. Five colonies (Nos. 2, 6, 11, 15, and 16) show maximum daily rates between 1,250 and 1,400, while the similar rates of six (Nos. 1, 5, 9, 10, 12 and 13) were between 1,000 and 1,250. In 1921 the maximum daily average of each of the 1919 queens is below 1,000. The time relations between maximum brood-rearing activity and nectar flows or pollen yields have already been discussed for each colony. The queen in colony No. 12, even in September, shortly after first beginning to lay, attained a daily average of 905. In the colonies studied in 1920 an egg-laying rate of over 1,500 per day was attained in only one colony, the queen in colony A averaging 1,528 eggs per day for one 21-day period. In colony D, in the same year, the queen attained a maximum daily rate of 1,468 eggs for one 21-day period, and the maximum daily egg-laying rates of the queens in colonies E, B, and C for any 21-day period were 1,223, 1,201, and 1,008, respectively. These daily averages may be compared with Du four’s (11) maximum daily average of 1,627 during any 21-day period. On the other hand, Baldensperger (/), in his estimates al- ready referred to, gives 2,600 as a daily average for a period of 23 days. It must be remembered, however, that Baldonspergcr’s method is not adapted for strictly accurate scientific results. Of the 36 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE daily egg-laying rates found by Brlinnicb (6) even the highest is slightly below 2,000. In fact, he states his belief that a daily rate of 2,000 eggs has never been exceeded in any of his colonies. At the end of his article already referred to, Dufour (11) makes a statement which applies with equal effect to the colonies used in this work. Although he recognizes that the egg-laying rates which he publishes are only averages and, as such were undoubtedly exceeded at times, he justly asserts that the results of his work do not warrant the assumption that any such daily egg-laying rates as 3,000 or more had ever been reached in any of the colonies used in his experiments. Since the daily egg-laying average for any season is far below the daily egg-laying average for any particular number of days within the maximum of that season, if is readily seen that the remarkably high rates of egg laying over short periods, so often published in beekeeping literature, can not be used as the daily averages for an entire season. CONCLUSIONS In this work no special effort has been made to modify the time of the various phases of brood rearing or to increase their intensity other than to provide adequate stores and ample room at most times; the colonies were therefore in much the same condition as might be found in an average apiary. The following conclusions may be drawn from the records of brood rearing presented here: The number of bees in the colony at the beginning of brood rearing in the spring, the ability of the queen, the abundance of stores, the suitability of the combs and proper insulation are the most important factors within the control of the beekeeper which, determine the amount of brood reared by a colony. The seasonal brood cycle in any region is marked by certain definite phases — the initial expansion, the major period, and the final con- traction. These tend to remain constant from year to year, their normal occurrence and magnitude being determined to a large degree by local weather conditions and by the local honey flows and pollen yields. A strong colony tends to retain its strength from year to year, other things being equal. x A queen at times transfers her egg-laying activity from one hive body to another, without any appreciable diminution in her rate of egg-laying if the combs are good. The possibility of young bees occurring to such an excess as to be conducive to swarming is reduced if the queen has ready access to another hive body, in case egg laying in the one already occupied becomes restricted through incoming pollen, nectar, or brood-rearing activity. Every colony used in 1921 shows a migration of the queen from one hive body to another, from which it may be inferred that if only one hive body had been available the amount of brood reared would have been reduced. There is a decided tendency for the brood area of the colony to be confined to adjacent combs in one or more hive bodies in such a way as to maintain the brood area in compact form. Although it can not be concluded from this investigation that the use of old queens is always disastrous, the records show that their use is accompanied with risk. THE BROOD-REARING CYCLE OF THE HONEYBEE 37 Prolonged inclement weather retards brood rearing in the spring, although a strong colony may be able to maintain its rate through unfaYorable cold weather of onty a few days’ duration, even though it is not packed. During the latter part of the active season the beekeeper may make important preparations for the next year’s honey crop by providing any of the factors which are necessary for the unrestricted increase of brood rearing during the period of initial expansion. Among the most important of these for every colony are good stores in sufficient quantity to last until incoming nectar suffices in the next active season, a good queen, and an abundance of worker comb. Although some of these conditions might be provided in the spring, any postponement is dangerous. Such preparations during the preceding active season, together with any others necessary for good wintering, may be expected to result in increased brood rearing so early in spring as to have the largest field force of the season avail- able during the nectar flow instead of after it has passed. LITERATURE CITED (1) Baldensperger, Ph. J. 1895. How many eggs does a queen average per day during the year? In Gleanings Bee Cult., vol. 23, pp. 950-951. (2) Baldridge, M. M. 1861. Fertility of the queen. In Amer. Bee Jour., vol. 1, pp. 109-110. (3) Berlepsch, Baron August von. 1860. Die Biene und die Bienenzucht. 475 pp. (4) Brunnich, Karl. 1912. Brutmessungen. In Schweiz. Bienen-Ztg., bd. 35, No. 7, pp. 257-261, 1 fig. (5) 1919. Das Messen der Brutflachen im Bienenstocke. In Schweiz. Bienen-Ztg., n. F., bd. 42, Nos. 6, 8, pp. 225-227, 283-288, 2 figs. (6) 1922. Graphische Darstellung der Legetatigkeit einer Bienenkonigin. In Arch. Bienenkunde, bd. 4, No. 4, pp. 137-147, 2 figs. (7) Demuth, George S. 1921. Swarm control. U. S. Dept, of Agr., Farmers’ Bui. 1198, 47 pp. (8) Desborough, J. G. 1852. On the duration of life in the queen, drone, and worker of the honeybee; to which are added observations on the practical importance of this knowledge in deciding whether to preserve stocks or swarms; being the prize essay of the Entomological Society of London for 1852. In Trans. Ent. Soc. London, 2d ser., vol. 2, pp. 145-171. (9) 1855. Observations on the honeybee, in continuation of the prize essay of the Entomological Society for the year 1852. In Trans. Ent. Soc. London, 2d ser., vol. 3, pp. 187-196. GO) 1868. Observations on the duration of life in the honey bee. In Trans. Ent. Soc. London, 3d ser., vol. 6, pp. 225-230. (11) Dufouh, Leon. 1901. Recherches sur la ponte do la rcine. In Annuaire do la F6d6ra- tion des Soci6t6s Frangaise d’ Apiculture, 10,no Session, pp. 18-34, 2 figs. (12) Morgenthaler; Otto. 1923. Eiriiges fiber die Krankheiten dcr erwaclisenen Bienen. In Schweiz. Bienen-Ztg., n. F., bd. 46, Nos. 1, 2, pp. 22-28, 81-85, 2 figs. (13) Reaumur, RfiN6 Antoine. 1740. Mdmoires pour servir i\ l’Histoire des Insectes, vol. 5, Paris. (14) Swammerdam, John. 1758. The book of nature, or the history of insects. (English trails, from the original Dutch and Latin editions. London.) 38 BULLETIN' 1349, U. S. DEPARTMENT OF' AGRICULTURE TABLES Table 1. — Record of seeded brood in colony No. 1 during the season of 1921, by vjeeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 16 4, 871 May 9 11, 090 July 5 10, 786 Aug. 29 5, 484 21 7,858 16 11,601 11 10, 134 Sept. 6 4, 655 28 9, 713 23 11, 596 18 9, 152 12 4, 563 Apr. 4 10, 100 31 12, 322 25 7,319 19 4, 012 11 11, 260 June 6 11, 256 Aug. 1 6,939 26 5,410 18 10, 140 13 10, 731 8 6, 139 Oct. 3 ' 4,485 25 9, 500 20 10, 970 15 5, 898 10 1, 772 May 2 10, 412 27 11, 148 22 6, 175 17 401 Table 2. — Record of sealed brood in colony No. 2 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 16 5,821 May 11 13, 551 July 6 9, 489 Aug. 31 6,543 23 11,438 18 13, 178 13 9, 853 Sept. 7 4, 908 31 12, 212 25 12, 873 20 13, 613 14 4, 920 Apr. 6 13, 366 June 1 13, 352 27 11,849 21 7,196 13 13, 067 8 14, 068 Aug. 3 10, 271 28 6,008 20 9,867 15 14, 693 10 9,587 Oct. 5 2,156 27 10, 822 22 15, 399 17 7,649 12 569 May 4 12, 814 29 15, 787 24 7, 591 19 473 Table 3. — Record of sealed brood in colony No. 3 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 17 4, 355 May 19 10, 821 July 14 5, 626 Sept. 8 1,098 25 8, 795 26 4,711 21 5,013 15 5,843 31 10, 414 June 2 3, 184 28 4, 432 22 9,582 Apr. 7 11,339 9 4, 928 Aug. 4 3, 272 29 4, 940 14 9, 736 16 6, 297 11 2, 214 Oct. 6 2,756 21 8, 667 23 6,804 18 1,832 13 1, 859 28 10, 543 30 6, 794 25 1 1, 639 20 708 May 5 11, 743 July 7 6, 177 Sept. 1 1,472 26 365 13 11, 614 1 Supersedure. Table 4. — Record of sealed brood in colony No. J+ during the season of 1921t by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 16 5, 012 May 10 16, 982 July 5 13, 659 Aug. 30 8, 898 23 9, 024 17 17, 859 12 11,741 Sept. 6 9, 448 29 11, 186 24 17, 155 19 11,084 13 8, 015 Apr. 5 12, 327 31 10, 955 26 9, 904 20 7, 862 12 14, 781 June 7 13, 079 Aug. 2 8, 204 27 9, 433 19 15, 402 14 12, 408 9 7, 840 Oct. 4 7,644 26 14, 072 21 13, 674 16 7, 459 11 3, 973 May 3 15,028 28 14, 588 23 7,871 18 1, 143 THE BROOD-REARING CYCLE OF THE HONEYBEE 39 Table 5. — Record of sealed brood in colony No. 5 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 4, 187 May 11 8, 134 July 6 11, 502 Aug. 31 5, 043 23 4,919 18 8, 763 13 10, 224 Sept. 7 1, 640 30 3, 949 25 11,047 20 8, 669 14 2, 121 Apr. 6 3, 696 June 1 12, 277 27 6, 168 21 3,300 13 4, 662 8 12, 830 Aug. 3 4, 571 28 3, 65i 20 4, 940 15 12, 567 10 3, 803 Oct. 5 1, 372 27 5, 765 22 13, 062 17 4, 187 12 May 6 8, 949 29 12, 405 24 7, 267 Table 6. — -Record of sealed brood in colony No. 6 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 5, 702 May 9 14, 829 July 5 13, 313 Aug. 29 6,029 21 7,593 16 15, 592 11 11,870 Sept. 6 4, 856 28 9,415 23 16,211 18 12, 679 12 6, 686 Apr. 4 9, 120 31 16, 401 25 12, 554 19 7,818 11 11, 536 June 6 14, 130 Aug. 1 11, 048 26 7, 871 18 12, 976 13 11, 893 8 9, 369 Oct. 3 4, 694 25 12, 772 20 13, 304 15 8, 296 10 945 May 2 14, 497 27 14, 690 22 7, 584 Table 7.— Record of sealed brood in colony No. 7 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 17 4,914 May 10 9, 923 July 12 2,496 Sept. 13 9, 769 22 8, 308 17 10, 319 19 2,015 20 10, 849 29 10, 332 24 9, 232 26 1,253 27 5, 888 Apr. 5 10, 904 June 1 7,558 Aug. 2 996 Oct. 4 4,492 12 11, 149 7 7, 348 9 659 11 3,066 19 10, 050 14 6, 977 16 454 18 669 26 10, 531 21 5, 436 23 338 May 3 10, 887 28 3, 650 30 i 127 6 10, 417 July 5 2, 354 Sept. 6 2, 873 1 Supersedure. Table 8. — Record of sealed brood in colony No. 8 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 16 4,241 May 19 8, 768 July 21 6,411 Sept. 22 1,014 25 7, 092 26 8,070 28 6, 154 29 4,380 31 7, 650 June 2 7, 591 Aug. 4 5, 856 Oct. 6 4, 147 Apr. 7 8, 504 9 7,669 11 5, 390 13 5 1, 152 14 8, 542 16 6, 402 18 5,614 20 m 21 7,623 23 6, 338 25 1 2, 685 26 520 28 8, 091 30 6,711 Sept. 1 (2) Nov. 3 2, 075 May 5 8,825 July 7 6,410 8 (3) 10 1, 136 14 9,088 14 6, 507 15 C) 1 Supersedure. 1 Virgin lost. 3 Rcqueened. 4 New queen laying. 1 Queenless, requconed. 0 New queen laying. 40 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE Table 9. — Record of sealed brood in colony No. 9 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 17 6, 397 May 9 11,245 July 5 11,375 Aug. 29 1 8, 274' 21 8,337 16 11,421 11 10, 789 Sept. 6 5,851 28 8, 686 23 11,670 18 10, 331 12 7, 842 Apr. 4 8, 443 31 11,638 25 10, 199 19 9, 004 11 10, 297 June 6 11, 408 Aug. 1 10, 549 26 8, 703 18 10. 951 13 10, 620 8 10,015 Oct. 3 5,831 25 11, 536 20 10, 744 15 9, 854 10 2, 350 May 2 12, 223 27 11, 521 22 9, 961 17 197 1 Supersedure. Table 10. — Record of sealed brood in colony No. 10 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 23 8,196 May 18 12, 352 July 13 10, 878 Sept. 7 6, 936 31 10, 218 25 12, 119 20 9, 645 14 6, 603 Apr. 6 10, 985 June 2 11,276 27 8, 661 21 8, 241 13 12, 392 8 10,717 Aug. 3 4, 191 28 8,675 21 12, 886 15 11, 294 10 (*) Oct. 5 5,677 27 12, 103 22 11,758 17 12 3, 165 May 4 11, 434 29 11,745 24 2, 188 19 907 13 12,242 July 6 11, 376 31 6, 102 25 695 1 Natural requeening. Table 11. — Record of sealed brood in colony No. 11 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 17 6, 475 May 14 15, 580 July 7 12, 928 Sept. 1 7, 281 25 12, 283 19 15,414 14 11, 840 8 4, 388 31 13, 788 26 16, 227 21 10, 847 15 5, 294 Apr. 7 16, 134 June 2 14, 678 28 10, 914 22 7,699 14 15, 712 9 10, 855 Aug. 4 10, 658 29 7,804 21 12, 837 16 14,916 11 9, 406 Oct. 6 4, 661 28 12, 388 23 15, 205 18 9, 687 13 1, 231 May 5 14, 585 30 15, 054 25 9, 731 Table 12. — Record of sealed brood in colony No. 12 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 5, 904 May 17 9, 930 July 19 9,213 Sept. 20 5,557 22 7,417 24 11, 651 26 8,289 27 (2) 29 9, 328 June 1 12, 095 Aug. 2 4, 675 Oct. 4 Apr. 5 11,017 7 11, 156 9 2, 768 11 316 12 12,913 14 10, 276 16 2,042 18 657 19 13, 229 21 10, 335 23 1,202 26 497 26 12, 855 28 10, 500 30 1372 Nov. 1 162 May 5 12, 682 July 5 13, 520 Sept. 6 5, 295 10 10, 718 12 9, 897 13 10, 871 1 Supersedure. 2 Natural requeening. THE BROOD-REARING CYCLE OF THE HONEYBEE 41 Table 13. — Record of sealed brood in colony No. IS during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 3, 404 May 13 10, 783 July 8 12, 680 Sept. 2 8, 564 25 6, 728 20 12, 684 15 11, 681 9 8,313 Apr. 1 8, 291 27 13, 220 22 11, 605 16 9, 511 8 9,444 June 3 13, 256 29 9, 871 23 10, 134 15 10, 533 10 14, 752 Aug. 5 4,648 30 6,891 22 9, 724 17 12, 234 12 0) Oct. 7 3, 655 29 8, 705 24 12, 732 19 14 1, 431 May 6 9, 768 July 1 12, 745 26 4, 141 21 379 i Natural requeening. Table 14. — Record of sealed brood in colony No. 14 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 9,101 May 13 12, 018 July 8 16, 098 Sept. 2 11, 404 25 12,412 20 14, 106 15 15, 463 9 7, 597 Apr. 1 11, 732 27 16, 479 22 14, 281 16 . 7, 404 8 12, 856 June 3 17, 175 29 11, 889 23 9, 436 15 12, 527 10 17, 414 Aug. 5 12, 299 30 5, 180 22 9, 995 17 18, 162 12 12, 929 Oct. 7 1,757 29 9, 545 24 18, 151 19 13, 126 14 963 May 6 12, 012 July 1 16, 779 26 13, 446 21 192 Table 15. — Record of sealed brood in colony No. 15 during the season of 1921, by weeks Date Scaled cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 6,514 May 13 8, 425 July 8 13, 949 Sept. 2 6, 976 25 8, 225 20 12, 957 15 12, 017 9 2, 489 Apr. 1 8, 469 27 15, 608 22 12, 293 16 4, 311 8 11, 605 June 3 15, 545 29 10, 907 23 8, 240 15 9, 938 10 15, 555 Aug. 5 8, 969 30 2, 929 22 7, 794 17 15, 625 12 7, 846 29 8, 531 24 16, 080 19 9,119 May 6 11, 214 July 1 15,239 26 10, 467 Table 16. — Record of sealed brood in colony No. 16 during the season of 1921, by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 18 6, 654 May 13 8,719 July 8 12, 686 Sept. 2 (') 25 12, 034 20 13, 043 15 8, 423 9 3,586 Apr. 1 13, 778 27 13, 974 22 10, 485 16 8, 555 8 12, 259 June 3 12, 914 29 8,914 23 6, 304 15 12, 949 10 14, 383 Aug. 5 8, 036 30 2, 987 22 5, 461 17 15, 845 12 6, 744 Oct. 7 1,701 29 3, 632 24 13, 397 19 3, 449 14 131 May 6 6,009 July 1 11,115 26 48 * Natural requeening. 42 BULLETIN 1949, U. S. DEPARTMENT OE AGRICULTURE Table 17. — Record of sealed brood in colony No. 4 during the season of 1922 , by weeks Date Sealed cells Date Sealed cells Date Sealed cells Date Sealed cells Mar. 10 3, 190 May 12 18, 694 July 14 10, 494 Sept. 15 8, 519 17 5,521 19 18, 661 21 9,597 22 8, 395 24 6, 125 26 17, 301 28 8, 703 29 8,870 31 6. 435 June 2 16, 204 Aug. 4 8,321 Oct. 6 6, 102 Apr. 7 9, 831 9 14, 457 11 8, 523 13 3, 756 14 14, 372 16 13, 658 18 8,872 20 2,986 21 17, 096 23 14, 154 25 10, 177 27 1,183 28 18,755 30 13, 942 Sept. 1 10, 548 May 5 19, 049 July 7 12,614 8 9, 649 Table 18. — Record of brood in colony A during the season of 1920, by weeks Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Mar. 15 2, 842 May 26 32, 092 Aug. 4 15, 343 Oct. 13 0 22 3, 206 June 2 30, 466 11 14, 472 20 2,065 . 29 7,748 9 28, 530 18 14, 253 27 3,493 Apr. 6 15, 686 16 23,886 25 0 Nov. 3 3, 519 14 11, 101 24 20, 611 Sept. 2 12, 562 10 3, 273 21 16, 622 30 19, 153 9 10,311 18 1,601 28 21, 160 July 8 19, 713 15 8, 939 24 860 May 3 25, 895 14 19, 034 22 9,585 12 30, 421 21 18,347 29 6, 489 19 30,258 28 16, 527 Oct. 6 3,340 1 No record taken. 2 Natural requeening. Table 19. — Record of brood in colony B during the season of 1920, by weeks Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Mar. 10 1,894 May 13 16, 656 July 15 17, 774 Sept. 16 9, 000 15 1,338 20 19, 683 22 17, 799 23 10,703 22 1,542 27 24,370 29 16, 358 30 12, 341 29 5,060 June 3 25,231 Aug. 5 1A 102 Oct. 7 6, 986 Apr. 7 8, 239 10 23, 314 12 12, 894 14 3, 417 15 6,983 17 23,349 19 12,835 21 1, 275 82 9, 093 25 25, 255 26 0 28 864 29 12, 595 July l 23, 927 Sept. 3 10, 452 Nov. 4 331 May 4 15, 461 9 18, 884 10 7,968 1 No record taken. Table 20. — Record of brood in colony C during the season of 1920, by weeks Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Mar. 17 2,609 May 17 18, 051 July 19 16, 873 Sept. 20 9, 866 24 3,497 24 20,194 26 16, 090 27 8, 572 31 7,159 29 21, 177 Aug. 2 15, 422 Oct. 4 7,613 4, 394 Apr. 7 8, 261 8,692 June 7 19, 536 9 15, 428 11 13 14 19, 150 16 16,446 18 1, 995 19 8,655 22 18, 678 23 0 25 824 26 May 5 10 12, 024 17,894 16, 626 28 July 6 12 19, 346 19, 198 18, 331 31 Sept. 7 13 14, 031 10, 525 9, 761 Nov. 1 155 1 No record taken. THE BROOD-BEARING CYCLE OF THE HONEYBEE 43 Table 21. — Record of brood in colony D during the season of 1920, by weeks Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Mar. 17 3, 397 May 18 23, 382 July 20 20, 910 Sept. 21 11,519 24 4, 987 25 29,411 27 18, 022 28 15,016 31 10, 222 June 1 30, 840 Aug. 3 19, 347 Oct. 5 10, 779 Apr. 7 13, 374 8 29, Q59 10 18, 631 12 6, 976 13 11,466 15 28,019 17 19, 189 19 4, 650 20 12, 564 23 26, 556 24 0) 26 3, 223 27 18, 691 29 26, 431 Sept. 1 2 7, 536 Nov. 2 633 May 6 24,788 July 7 21, 777 8 3, 234 11 25, 157 13 20, 822 14 7, 120 1 No record taken. 2 Supersedure. Table 22. — Record of brood in colony E during the season of 1920, by weeks Date Total o feggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Date Total of eggs, larvae, and sealed cells Mar. 17 1,037 May 21 19,546 July 23 13, 190 Sept. 24 8, 384 24 1,567 28 22, 398 30 10, 686 Oct. 1 8,918 31 4, 897 June 4 23, 523 Aug. 6 11, 809 8 6, 182 Apr. 8 6, 078 11 25, 685 13 12, 826 15 4, 652 15 (>) 18 22, 250 21 12, 519 22 2, 877 23 6, 458 26 21, 706 28 (2) 29 1,304 30 11, 255 July 2 22, 469 Sept. 4 12, 251 May 7 14, 821 10 22, 077 11 9, 330 14 15, 537 16 18, 077 17 8,426 e_ 1 Requeening. 2 No record taken. 44 * BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE GRAPHS Figs. 1 to 4.— Curves showing amount of sealed brood found weekly during the season of 1921 in colonies Nos. 1 to 4 THE BROOD-REARING CYCLE OF THE HONEYBEE Figs. 5 to 8.— Curves showing amount of sealed brood found weekly during the soason of 1021 in colonies Nos. 5 to 8 46 BULLETIN 1349, U. S. DEPARTMENT OP AGRICULTURE Figs. 9 to 12— Curves showing amount of sealed brood found weekly during the season of 1921 in colonies Nos. 9 to 12 THE BBOOD-REAEING CYCLE OF THE HONEYBEE 47 Fios. 13 to 10. — Curves showing amount of sealed brood found weekly during the season of 1921 in colonies Nos. 13 to IS 48 BULLETIN 1349, U. S. DEPARTMENT OE AGRICULTURE Fig. 17. — Curve showing amount of sealed Fig. 18.— Curves showing amount of sealed brood found weekly during the season of 1922 brood found weekly during the seasons of in colony No. 4 1921 and 1922 in colony No. 4. The curve for 1921 is represented by a broken line; that for 1922 by an unbroken line Figs. 19 and 20. — Curves showing total eggs, unsealed larvae, and sealed brood found weekly during the season of 1920 in colonies A and B United States Department of Agriculture Department Bulletin 1349, "The Brood-Hearing Cycle of the Honeybee. 55 Cor rcctlon Slip Page 49, figs. 21 to 23. The numbers at the left of the horizontal lines in the graphs should be the same as in Figures 19 and 20, page 4S, namely, 40,000, 30,000, 20,000, 10,000 and 0. . THE BROOD-BEARING CYCLE OF THE HONEYBEE 49 I'igs. 21 to 23. — Curves showing total eggs, unsealed larvae, and sealed brood found weekly during the season of 1920 in colonies C, D, and E 50 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE Tig. 24. — Curves showing amount of sealed brood found weekly in the various hive bodies during the season of 1921 in colony No. 4. The heavy unbroken line represents the total of all hive bodies; the dotted line the first, the narrow unbroken line the second, and the narrow broken line the third hive body THE BROOD-REARING CYCLE OF THE HONEYBEE 51 Fig. 25.— Curves showing variation in sealed brood found weekly in the various hive bodies dur- ing the season of 1922 in colony No. 4. The heavy unbroken lino represents the total of all hivo bodies; the dotted line the first, the narrow unbroken lino the second, and the narrow broken line the third hive body 52 BULLETIN" 1349, U. S. DEPARTMENT OF AGRICULTURE Ml ii m § M4P.23 NYJP.3Q Y?PP. 5 AfP/S./Z /9P&J9 n U Ii h <11 it /9PB.Z6 AtPYJ AMY/O MyPY/7 MYTYS4 MY7Y3! pH III JUNE 7 JUNE/4 JUNE 2/ JUNE 23 JULY 3 JULY! 2 Fig. 26. — Diagrammatic representation of the amount of sealed brood on each side of each frame in the three hive bodies of colony No. 4 from March 16 to July 12, 1921 THE BROOD-REARING CYCLE OP THE HONEYBEE 53 tJULV /& k/(SLY26 S7UG.2 S9C/G.9 sGt/G./6 YfC/G. 23 Fig. 27.— Diagrammatic representation of the amount of sealed brood on each side of each_frame in the three hive bodies of colony No. 4 from July 19 to October 18, 1921 54 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE ^1 IIIBl llillll W'W Ikfelrilil MAR./O MAR./7 MAR. 24 MAR. 3/ ARR.7 APR./4 APR. 2/ APR. 26 MAYS MAY /2 MAY /9 MAY 26 M|| 1... 4\ it li « JURE 2 3URE 9 3UA/E/6 'JURE 23 30A/E30 3ULY 7 Fig. 28. — Diagrammatic representation of the amount of sealed brood on each side of each frame in the three hive bodies of colony No. 4 from March 10 to July 7, 1922 THE BROOD-BEARING CYCLE OF THE HONEYBEE 55 JULY /J- JULY 2/ JULY 23 Y?UG.* Jt s9UG./T Fig. 29.— Diagrammatic representation of the amount of sealed brood on each si< in the three hive bodies of colony No. 4 from July 14 to October 27, J IlH s4UG. /& SEPT 29 of each frame ORGANIZATION OF THE UNITED STATES DEPARTMENT OF AGRICULTURE September 8, 1925 Secretary of Agriculture W. M. Jardine. Assistant Secretary R. W. Dunlap. Director of Scientific Work Director of Regulatory Work Walter G. Campbell. Director of Extension Work C. W. Warburton. Director of Information Nelson Antrim Crawford. Director of Personnel and Business Admin- istration W. W. Stockberger. Solicitor R. W. Williams. Weather Bureau Charles F. Marvin, Chief. Bureau of Agricultural Economics Thomas P. Cooper, Chief. Bureau of Animal Industry John R. Mohler, Chief. Bureau of Plant Industry William A. Taylor, Chief. Forest Service . W. B. Greeley, Chief. Bureau of Chemistry C. A. Browne, Chief. Bureau of Soils Milton Whitney, Chief. Bureau of Entomology L. O. Howard, Chief. Bureau of Biological Survey E. W. Nelson, Chief. Bureau of Public Roads Thomas H. MacDonald, Chief. Bureau of Home Economics Louise Stanley, Chief. Bureau of Dairying C. W. Larson, Chief. Fixed Nitrogen Research Laboratory F. G. Cottrell, Director. Office of Experiment Stations E. W. Allen, Chief. Office of Cooperative Extension Work C. B. Smith, Chief. Library Claribel R. Barnett, Librarian. Federal Horticultural Board C. L. Marlatt, Chairman. Insecticide and Fungicide Board J. K. Haywood, Chairman. Packers and Stockyards Administration John T. Caine, in Charge. Grain Futures Administration J. W. T. Duvel, in Charge. This bulletin is a contribution from Bureau of Entomology L. O. Howard, Chief. Division of%Bee Culture Investigations J. I. Hambleton, Apiculturist, in Charge. 56 ADDITIONAL COPIES OF THIS PUBLICATION MAT BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY V