°)45 m UNITED STATES DEPARTMENT OF AGRICULTURE m BULLETIN No. 273 *^^|><J ^^*^^U L. O. HOWARD, Chief S^^^'^^^U IWashington, D. C. PROFESSIONAL PAPER August 24, 1915. DISPERSION OF GIPSY-MOTH LARV^ BY THE WIND/ By C. W. Collins, Entomological Assistant, Gipsy Moth and Brown-tail Moth Investigations. CONTENTS. I Previous investigations : Page. On dispersion 2 Of hairs on small larvae 3 I Studies of acuminate and vesicular hairs 4 jScope of investigations 6 JHatehing in relation to wind dispersion 7 I Longevity of entrapped first-stage larvae 8 I Screen experiments 8 [Dispersion experiments, New Hampshire hills 14 Altitude experiment 15 Observations on wind dispersion 17 Observations for reinfestation 17 Small larvae blown into ocean 18 Feeding prior to dispersion 19 Comparison of weather data 19 Spread of gipsy moth 20 Summary 21 Bibliography 22 INTRODUCTION. The territory infested by the gipsy moth {Porthetria dispar L.) is [gradually increasing in area, and since isolated colonies are being found each year in woodlands and orchards remote from lines of kravel, the question as to then origin has become more urgent. Many agencies assist in the transportation of the moth in its various stages, but as previous experiments proved that small caterpillars are borne by the wind, it seemed desirable to ascertain to what extent [this occurs and the maximum distance they may be blown under I field conditions. Much depends upon this factor in diffusion, when devismg methods of preventuig the spread of this pest, and as a I practical result of previous experiments tanglefoot has been used extensively in colonies located in the outside infested towns. Ti'ees are banded with tanglefoot to prevent the small larvae from ascending to the tops, and thus the chances that these larvae will be blown long distances are decreased. 1 An extended report on one of the important phases of gipsy moth control work, conducted under the direction of the Biu-eau of Entomology, L. O, Howard, Chief, and in cooperation with the various States affected. 968io°-Buii. 273-15 — 1 Mooograph 2 BULLETIN 273, U. S. DEPARTMENT OF AGEICULTURE. To secm-e further information, the writer, under the direction of Mr. A. F. Burgess, commenced a series of experiments in the spring of 1913 and continued them throughout the following year.^ PREVIOUS INVESTIGATIONS ON DISPERSION OF THE GIPSY MOTH. Many mvestigations of various sources of spread of this insect were conducted during the early nineties by Forbush and Femald and pubhshed by the Massachusetts State Board of Agriculture in 1896. It was found that frequent teamuig in and out of the origmal infested areas (Medford, Boston, and towns closely surroundmg) by vehicles transportmg larvse that dropped from infested shade and roadside trees was largely responsible for the general local spread during those years. It was also suggested that the larvae were occasionally trans- ported by the wind to distances of perhaps 100 yards, but no further observations were made at that time to substantiate this theory. Numerous ways and means of local spread, such as by birds, animals, and man, are treated m this valuable report. In the sprmg of 1910 Mi\ A. F. Burgess began a series of experiments on dispersion of the gipsy moth the account of which is pubhshed as Bulletin 119 of the Bureau of Entomology. A few small experiments were made in the laboratory by releasmg silk-suspended first-stage gipsy-moth larvas in front of an electric fan, and it was found that they soared to a distance of 20 to 30 feet in the air before ffdling. This led to more extensive experiments conducted under out-of-door conditions. Screens constructed of poultry wire and covered with tanglefoot were stationed at given distances from sources of infesta- tion and watched for catches of larvae. On a small screen stationed on the marshes near Lynn, Mass., newly hatched larvae were caught that had been carried 1,833 feet by a west wuid at a velocity of 7 to 19 miles per hour. Numerous records were compiled to show the relation of tempera- ture, wind velocity, and wuid direction to the general trend of spread as it has progressed from the original center, Medford, Mass. It was clearly demonstrated that the prevalent southwest winds following the hatching period of the eggs each year has resulted in the more rapid dispersion of the moth to the northeast than m any other direc- tion. Large woodland areas were scouted m various towns, the data of which were compiled and show the extent to which tlie wuid is responsible for the remote and numerous colonies located. The automobile is also included in this report, as a very important factor in carrying the caterpillars both locally and for long distances. 1 The writer is indebted to Mr. L. H. Worthley and Dr. J. W. Chapman for cooperation and helpful suggestions; to Messrs. C. E. Hood, F. W. Graham, and Harry lilaisdell for the collection of field data; to Mr. H. A. Preston for the preparation of photographic illustration-:, and to others oi the laboratory iorce who contributed in various ways. D. oi! D. SEP ^i 1916 (5 V^ CS^^ DISPEESION OF GIPSY-MOTH LARV^ BY THE WIND. &^^ FORMER INVESTIGATIONS OF HAIRS ON SMALL LARV^. The bodies of first-stage laiTse (PI. I, fig. 1) are pi'ovided both, dorsally and laterally with two types of hairs, or setse, namely, short, smooth hairs bearing a vesicle about the middle and very long, slender ones which are covered with spinules. Wachtl and Komauth in 1893 first described the aerostatic seta? fomid on the first-stage larvse of P. dis'par L. and Lymantria rnonacJia L. and designated the bdHoon-shaped swellings occurring on these setae as aerophores. They suggest that these aerophores assist ui the dissemination of the young caterpillars through the air. Prof. Cholodkovsky in 1894 made some investigations and found that the swelhngs or vesicles shrank in dead larvae. His discovery tended to weaken the theory that these swellings contam air and to suggest that they may contain a fluid which will naturally dry up after death. He also fomid that the sweUings remained for months in alcohol as full and rounded as m hvuig lai-vae, and if the preparation was allowed to dry on a shde the aerophores quickly shriveled. He therefore concluded that they did not contain air but a fluid which was probably poisonous and served as a protection against insectivorous birds. Igenitzky, a student of Cholodkovsky, in 1897 further studied the glands that give rise to the hairs and verified the findings of the former investigator, who proposed to call the sweUings " toxophores. " He further states that the role of rendering the larvae more buoyant may better be ascribed to the long thin hairs which resemble the pappus of some plant seeds. Prof. K. Escherich in 1912 published a resume of former investiga- tions into the function of the hairs and cites later work by Wachtl and Kornauth in 1907 in which they cling to their former theory of aero- phores. The latter found that the vesicles did not contain liquid, as no reaction was noted by immersion in litmus, rosolic acid, or phenol- phthalein, indicating that they contained neither, alkali nor acid. They did not shrink in alcohol, glycerin, or acetic acid, or excite any capillary action; hence the mvestigators concluded that nothmg but air could be contained in them. Prof. Escherich lays stress on the air refraction noted in connection with these balloons when immersed ia glycerin and viewed under the microscope. T. C. Shcherbakov in 1914 published observations on the gipsy moth in which he deals at length with the function of the hairs on the first and second stage larvae, and parts of his paper have been trans- lated by Mr. J. Kotinsky of the Bureau of Entomology. Shcher- bakov says that the aerophores are not filled with air or gas and that their connection with the glandular cells would indicate that these vesicles are probably filled with a poisonous secretion. Their exceed- 4 BULLETIN 273^ U. S. DEPARTMENT OF AGRICULTUEE. ingly small size, he contends, makes them physically inadequate to sustain the caterpillar in the air. On the other hand, the very numer- ous, very long hairs which cover the larva, in addition to the wonder- ful amount of silk it is capable of spinning, point to the general morpho-physiological structure as the factor enabling it to soar. He compares this ability of the caterpillar with the meaning of soaring as this term is used by V. N. Chitrovo in an epoch-making botanical work on the study of weeds.^ Soaring is the capability of a body passively to participate in an air current, as if constituting part of it, arising from special characters in the structure of the body. The analogy in function between the wmd-borne seed and caterpillar is complete. Soaring is determined by the coefficient calculated by the following formula: ^_ /(cm-) w(gr) where C is the soaring coefficient; /, the surface area (in square cm.); and w, the weight (in grams) of the caterpillar. The coefficient of a bare, freslily hatched first-stage caterpillar is 514.285 and Shcher- bakov says that it would be more than six times as great if the area of the hairs were included. Compared with the Chitrovo soaring table of seeds, the coefficient for the first and second stages of the caterpillar places them among the greatest seed soarers. In a slight wind these caterpillars are capable of soarmg. The distance they can cover depends on the topography of the country, the character and abundance of vegeta- tion, the mobility of the caterpillar which in turn depends upon the temperature of the atmosphere, and the availability of foods. The soaring coefficient diminishes sharply for caterpillars in the second stage. This investigator also states that G. G. Jacobson once, while trav- eling, observed caterpillars flying in masses toward him, despite the fact that he was at least 17 miles away from the nearest forest. STUDIES OF STRUCTURE AND FUNCTION OF ACUMINATE AND VESICU- LAR HAIRS. Owing to the differences of opinion existing as to the structure and function of the hairs covering the bodies of first-stage larvse, the \ATiter and Mr. C. E. Hood, with the assistance of Dr. J. W. Chapman, made some tests during the winter of 1914-15. The aim of these experiments was to study both the long acuminate hairs and the short ones bearing vesicles. As the results differ somewhat from those of 1 Atlas von Samen und Friichten der Feldunkraiiter aus Mittelmssland. Bui. fiir angewandte Botanlk, wissenschaftliches Organ des Bureau fiir angewandte Botanik (Monatsschrift), Jahrg. 7, no. 3, 118 p., 16 fig., 13 pi., March, St. Petersburg, 1914. Bui. 273, U. S. Dept. of Agriculture Plate Fig. 1. -First-Stage Larva Showing Two Kinds of Hairs. (From Burgess.) Fig. 2.-a, Tubercle from First-Stage Larva Showing Hairs; b, Vesicular Hairs; c, Acuminate Hairs. All Much Enlarged. 'Original.) GIPSY-MOTH CATERPILLAR SHOWING VESICULAR AND ACUMINATE HAIRS. DISPERSION OF GIPSY-MOTH LAEV^ BY THE WIND. 5 previous investigators in that the swellings on the short hairs were found to contain a liquid not yet proven to have toxic properties, it seems advisable that they should hereafter be called vesicular hahs. (PI. I, fig. 2, &.) The term "aerophore" previously applied to the swellings on the short hairs is misleading, as is also the tenn "toxo- phore" suggested by Cholodkovsky. The term "acuminate" stdl applies to the very long slender hahs. (PL I, fig. 2, c.) After first-stage larvae die the sweUmgs on the hairs usually col- lapse and both air and liquid are present in each tjrpe of hair and can be seen under a high-power microscope. A living caterpillar exam- ined in glycerin shows air bubbles occasionally in the hahs and swellings which sometimes extend above and below the latter. The air globules and liquid columns alternate m hairs of dead larvee, and the difference between the air or colorless liquid present is evident from the typical air refractions in the globular form that air always takes in a liquid; also from the capillary attraction of the liquid contamed. Living caterpillars mounted in balsam and examined under a micro- scope show air bubbles in some of the hairs, but such hairs are excep- tions rather than the rule. Caterpillars that have been dead several days show the opposite phenomenon — both air and liquid are seen in them, but air predommates. Some larvae that had been dead for two years were examined and these contained small short columns of liquid, but air filled the greater portion of the hollow spaces of the hairs. Practically the same was noted m the hairs of an exuvium from a first-stage larva which had also been kept about two years. The vesicles on the hairs had buckled or collapsed and contained air, which was later practically all driven out by heating. Embryos were dissected from the chorion under balsam and glycerin and then examined in these media. No air globules or what could be termed air refractions were noted in either acuminate or vesicular hairs. Many of the vesicles are only partially distended in tliis stage. Embryos dissected from the chorion in air contained air globules and columns of fiquid alternating in a portion of both types of hairs — ^probably those ruptured during the dissection. After some experience it is easy with the aid of a high-power miscroscope to recognize the difference in refractions tlii'ough the walls of hairs full of colorless liquid or containing globides of air and columns of fiquid alternating. The typical fiquid refractions are always seen in uninjured vesicular and acuminate hairs. Only rarely are there air refractions noted in either type on living larvae. A decided change in aspect is noted in dead larvae after the vesicle begins to collapse, at which time air finds its way into the hairs. A number of fiving larvae were placed under a beU jar wliich con- tained a small amount of concentrated sidphuric acid. After six hours many of the larvae were dead and upon examination a large 6 BULLETIN" 213, U. S. DEPARTMENT OF AGRICULTUEE. portion of the vesicles of the haii-s had buckled or partly collapsed. Larvae still living at this time presented the same appearance to a slight extent. As a check on this experiment, larvse were killed with cliloroform and after 24 hom-s showed a smaller pro2:)ortion of buckled or collapsed vesicles, indicating that sulphuric acid took up moisture from the porous hairs. Sections were made of the larvaG and cells at the base of the hairs studied, but there coidd be found no indication of the presence of glandular cells. A large trichogen, however, is present at the base of each hair. This would signify that the liquid inclosed in the hairs is not a toxin, as Cholodkovsky suggests, but a mere colorless mobile hquid secreted diu-ing the formation of the hairs. The phenolphthalein and litmus-solution tests were also tried, with nega- tive results, further indicating a neutral liquid. The peculiar shape of the vesicular hairs, however, suggests that they may at one time have had a poisonous function but that it has been lost in the present generations. The results of the experunents and observations indicate that both the normal vesicular and acmiiinate haire are idled with a colorless liquid; that the hairs are hollow throughout, and that the vesicles contain the same media as the remaining portion of that type. A few globules of air were seen in hairs of living larvae wliich were no ted as exceptions. Air partially replaces the liquid after death, following which period the vesicles collapse. It is therefore probable, as Ige- nitzky and Shcherbakov have indicated, that the acuminate hairs play the greater role in making the larvse more buoyant, as these are from four to six times as long as the vesicular hairs. SCOPE OF INVESTIGATIONS ON WIND DISPERSION. In that the main purpose of the following investigations was to secm*e data on the maximum distance and the extent to which small larvae are borne by the wind, it was necessary to find conditions where an abundance of larvae were present in close proximity to treeless areas. These conditions were best afforded along the beaches in Massachusetts and New Hampshire, where there are stretches of marehland from 1 to 2^ miles wide and many miles long. These marshlands are occasionally flooded with salt water and do not con- tain vegetation favorable to the development of gipsy-moth larvae. The areas selected for the experiments were to the east of the heavy infestations, thereby getting all the advantages previously known to accrue from the northwest, west, and southwest winds. The Isles of Shoals, off the coast of New Hampsliire, afforded ideal conditions in so far as their remoteness from the mainland was concerned and the fact that the country opposite has been infested since 1905. Locations were also selected in the liills of New Hampsliire near the DISPERSION OF GIPSY-MOTH LAKViE BY THE WIND. quarantine line for experiments intended to explain the sources of many infestations found under such conditions. Screens of poultry wire, to which tanglefoot was applied, were erected in three of the selected locations along the coast and on the Isles of Shoals (PI. II). Those used in the New Hampsliirc liill ex- periments were of a somewhat different type and are described xmder that head. The screens used along the coast (PI. Ill, fig. 1) were of ^-inch mesh poidtry wire stapled to 2 by 4 posts which were set in the ground and held in place at the top by guy wires and wooden braces. The posts were 12 feet high and two sections of wire each 75 feet long and 3 feet wide were fastened to these so that the top selvage was 12 and the lower 6 feet from the ground. The screens contained 450 square feet of wire and were built in three 25-foot sections, the middle section facing the west and the other two anghng from tlie ends about 45° to the eastward. This gave the greatest exposure at all times to the northwest, west, and south- west winds. A runway was attached to the posts at the bottom selvage of the wire to facihtate examinations. HATCHING OF EGGS IN RELATION TO WIND DISPERSION OF FIRST- STAGE LARV^. Close observations on the time of hatcliing of egg clusters in the field have been kept since 1912 by laboratory men stationed in different sections of Massachusetts and New Hampshire. The data consulted were collected by Messrs. E. A. Proctor in northeastern Massachusetts, J. V. Schaffner, jr., in Massachusetts south and south- west of Boston, and I. L. Bailey in south-central New Hampshire. Table I. — First, maximum, and last dates of hatching in 1912, 1913, and 1914- Year. Northeastern Massachusetts hatching. Southeastern Massachusetts hatching. South-central New Hamp- shire hatching. First. Maxi- mum. Last. First. Maxi- mum. Last. First. Maxi- mum. Last. 1912 1913 Mav 2 Apr. 25 May 11 May 11 May 5 May 15 May 22 May 27 May 28 Mav 1 Apr. 26 May 9 May 15 May 10 May 15 May 25 May 21 May 23 May 1 Apr. 29 May 11 May 14 May 8 May 16 May 23 May 14 1914 May 28 The first larvse were caught on the screen in northeastern Massa- chusetts May 9, 1913, and the last June 5, while the largest numbers were removed between May 14 and Jmie 1. During 1914 in the same section the first larva was caught about May 16, while the largest numbers were removed from May 20 to May 27 and the last June 6. One larva was caught in Hemiiker, N. H., as late as June 13, 1914. In a year of normal hatching, as in 1913, wind dispersion was noted about two weeks after first hatching and one week after hatch- 8 BULLETIN 273, U. S. DEPARTMENT OF AGRICULTURE. ing was completed. In 1914 there was an abnormally late spring, which shortened the hatchmg season, and wind dispersion was noted about one week after the first date and continued about the same period after the last was observed. The total period over which wmdspread may be expected is from 27 to 30 days m a normal spring, and from 18 to 20 days during a late spring. Ordinarily the maximum dispersion is effected durmg 10 to 15 days of high tempera- tures and favorable winds. LONGEVITY OF FIRST-STAGE LARV^ AFTER BECOMING ENTRAPPED IN TANGLEFOOT. In connection with the experiments conducted by using tanglefoot on screens to catch caterpillars blown by the wind, it became essential to know about how long they would live after becoming entangled. As a test, tanglefoot was smeared at various depths on heavy paper and 100 caterpillars dropped into these. Exammations were made at intervals from the time the experiment was started, and it was noted that after three hours practically aU were living; after six hours only two were attemptmg to move, and these were in a very thin smear. After eight hours 65 were dead, and 35 showed signs of life when disturbed. The following morning, after 24 hours, aU were dead except seven, and these were in very thin smears, which had little effect upon them. The length of life depended upon the density of the tanglefoot in which they were placed. If the caterpillars m attempting to free themselves became submerged below the spiracles, which often hap- pened in 3 to 5 hours, they died shortly after. As a result, cater- pillars lodged in globules of tanglefoot on the screens died in 4 to 6 hours, while those in very thin spots lived 24 or more. SCREEN EXPERIMENTS AT SALISBURY BEACH, MASS., IN 1913. April 24, 1913, a large screen (PL III, fig. 1), as described, was erected on the edge of the marsh area near the beach, on a site which was 1 mile distant from infested woodlands to the northwest, IJ miles to the west, and 2 J miles to the southwest. The area to the east, composed of sand dunes, contained small amounts of vegetation, on which gipsy moths could thrive, but this was properly scouted and cleaned for an experiment on reinf estation by the wind. The temperature, wmd direction, and wind velocity were taken each day at alternate hours from 9 a. m. to 5 p. m., and the screen was examined most days when the winds were fair for windspread from the west. Mr. F. W. Graham was in attendance at this screen and made many of tlie following observations and notes. No larvae were blown on the screen, which was one-fifth mile from the ocean, by wmds coming directly from the east. Larvae were Bui. 273, U. S. Dept of Agriculture. Plate II. MAP showing LOCATION «■ SCREEN TFtAPS where GIPSY MOTH LARVAE were eaii^t In 1013 ■-> 1014. Map Showing Location of Screen Traps where Gipsy-Moth Larv^ were Caught in 1913 and 1914. Bui. 273, U. S. Dept. of Agriculture. Plate III. . • A vj^a- 1 1--^^^ „.jBfc«» — .« mum*' ■RMMHMnva^ - •~ A . Fig. 1.— Screen at Salisbury Beach, Mass., where 266 Larv/e were Caught During May and June, 1913, and 96 in May, 1914. Showing Beach and Ocean in Background. (Original.) Fig. 2.— Screen at Plum Island, Mass., where 42 Larv/e were Caught in May, 1914. Newburyport in Background 2 Miles West— Nearest Infested Tree Growth in that Direction. (Original.) SCREEN TRAPS FOR WIND-BLOWN GIPSY-MOTH LARV/E ON MASSACHUSETTS COAST. DISPEESION OF GIPSY-MOTH LARV^ BY THE WIND. 9 borne, however, by northeast and southeast winds from beach infes- tations 1 mile or more distant. This screen was inspected in May by Dr. Howard and Dr. Paul Marchal. Table II. — Dates and numbers of newly hatched caterpillars caught on screen, ivith tem- peratures, direction of wind, and velocity of tvind, at Salisbury Beach, Mass., in 191S. Date and time removed. Number of caterpil- lars caught. Maxi- mum temper- ature. Minimum tempera- ture. Average between 9 a. m. and 5 p. m. Prevailing wind direc- tion. Wind veloci- ity, in miles per hour. May 9, p. m May 10 , May 11, a. m May 12 May 13 May It, a. m May 15, a. m. and p..m May 16, p. m '. . . May 17, p. m May IS May 19, p. m May 20, p. m May 21, p. ra May 22 May 2.3, p. m May 24 May 25 May 23, p. m May 27 May 28 May 23 May 30, p. m May 31, p. m Juiic 1, a. m. and p. m. June 2, p. m June 3, p. m June 4, p. m Junes, p. m Total (') (0 (') 0) (0 (0 (1) (}) (?) S.andSE... W. andSW. NW NW NW.andSE SW NW W. and SW. NE E. and NE.. 11 to 15 8to U 18 to 22 15 to 18 8 to 11 10 to 13 13 to 17 7 to 9 8 to 12 8 to 12 W. and NW NW.andW. SE E.toNw!!! NW. SW., and E. N. to E E . and SE . . NW to E . . . E n:: NW W.andNW. NE to NW . SE toSW... NW SW NW. toE... E.toS 18 to 23 15 to 20 8 to 12 6 to 8 2 to 5 7 to 10 7 to 11 6 to 10 8 to 12 8 to 11 19 to 23 17 to 21 8 to 12 16 to 20 16 to 20 15 to 19 12 to 16 7 to 11 1 No examination. - No records. It wiU be noted from Table II that 266 caterpillars were caught on 450 square feet of wire, and most of these on days when the winds blew from the northwest, west, or southwest, namely, May 14 to 19, May 15 to 17, May 19 to 25, May 20 to 30, etc., while fewer were caught when the wind was from the northeast and southeast. This may be explained by the light infestations 1 mile or more to the north- east and southeast; also it will be noted that winds from these direc- tions along the coast are accompanied by much lower temperatures, at which time the caterpillars spin less actively. Many of the small larvee were taken on the posts and crosspieces of the structure; thus of the 25 caught May 19, 18 had lodged in this manner and were living when removed. Others were swinging by threads on the east side of wire, where they had blown through the mesh from the west, the threads having caught in tanglefoot. 96810°— Bull. 273—15 2 10 BULLETIN 273, U. S. DEPARTMENT OF AGRICULTUKE. The first larva was caught May 9 and the last June 5, although exammations of later date were made. These dates indicate the range of the wind dispersion period for that section in 1913. The distance borne by the wmd was 1 mile or more. Another small experiment was conducted 200 feet from the above screen by attaching to the northwest, west, and southwest sides of a small stable 77 square feet of cotton cloth which was smeared with tanglefoot. This experiment was not started at the beginning of the season, but larvae were blown into the tanglefooted cloth be- tween May 20 and 31. The object in this experiment was to ascer- tain the comparative efficiency of tanglefooted wire screen and a solid tanglefooted surface as traps. CONTINUATION OF SCREEN EXPERIMENT AT SALISBURY BEACH, MASS., IN 1914. The same screen (PL III, fig. 1) as was erected in 1913 was cleaned and retanglefooted for the dispersion season of 1914. The infesta- tion in the woodlands to the westward, owing to the increase of Calosoma sycophanta L., Com.psilura concinnata Meig., and the wilt disease, was less intense during the latter year. The dispersion period was much later and consequently much shorter than in 1913. Examinations were begun about the time the hatching of eggs was first noted in the field. The first caterpillar was caught May 19 and the last May 31 in this location. The weather records were taken 1 mile south for this point at Plum Island, Mass. (PL II), by Mr. C. E. Hood, and were compared with those of the United States Weather Bureau at Boston, Mass., from which they differed very little. Table III.^ — Dates and numhers of newly hatched caterpillars caught on screen, with tem- perature, direction of wind, and velocity of wind, at Salisbury Beach, Mass., in 1914- Date and time removed. Number of cater- pillars cauglit. Maxi- mum tempera- ture. Mini- mum tempera- ture. Average between 8 a. ra. and 5 p. m. PrevaLUng wind direction. Wind veloc- ity, in miles per hour. May 19, a. m May 20, a. m May 21, a. m. and p. m May 22, a. m May 23, a. m. and p. m May 24, a. m. and p. m May 25, a. m May 2r), a. m May 27, a. m May 28, a. m May 29, a. m. and p. m May 30, a. m May 31, a. m. and p. m Total "F. °F. W. and NW... W. andNW... NK. toSE E.andSE W W sw SW.and W...- W.toS NW E.andSE SW W 8 to 12 2 to 6 7 to 11 2 to 10 10 to 23 11 to 16 12 to 21 7 to 20 4 to 12 11 to 20 9 to 17 7 to 9 7 to 14 DISPERSION OF GIPSY-MOTH LARV^ BY THE WIND. 11 In all 96 larvse were trapped during tlie season, most of which were removed on days when the wind blew from the west, southwest, or northwest. The largest catch was removed May 21, at wliich time the wind blew from the northeast and southeast, but it was evident that most of the larvae were caught May 20 in the afternoon, as the screen was not examined durmg that period. There was less source of spread from the northeast and southeast along the ocean shore than from the westward points. On May 21 the temperature ranged from 53° to 60° F., at which time there was reasonable activity among the caterpillars. One should not fail to note that spread was secured from all direc- tions from which the wind blew except from the east, and the in- festation at this point Was almost obliterated. No north winds were noted during the hourly observation periods. SCREEN EXPERIMENT CONDUCTED ON PLUM ISLAND, MASS., IN 1914. April 14, 1914, a screen (PL III, fig. 2) was erected on the edge of the marsh near the Merrimac River. It was of the same size and proportions as that erected 1 mile farther north, at Salisbury Beach. The slirubbery, composed mostly of beach plum, bayberry, wild rose, sumac, poplar, and wiUow, on the north end of the island, was ex- amined for egg clusters and these were creosoted. This practically obliterated the source of infestation to the eastward of the screen. The location of the screen (PL III, fig. 2) was such that it was 2 miles from infested woodlands across the marshes and Merrimac River to the northwest, west, and southwest. Frequent examina- tions made it possible to determine the direction from which all the caterpillars came. It is not supposed that all the caterpillars caught during the time west winds prevailed came from the nearest infesta- tions (2 miles), but in some instances 5 miles or more. There was, however, a small infestation on a few roadside willows 1 mile to the southwest, but these were so thoroughly cleaned and tanglefooted in the spring that no caterpillars could be found on them during the dispersion season. The woodland infestation was 1 mile bej^ond this, totaling 2 miles from the screen. Mr. C. E. Hood attended the screen, making three or four exami- nations for larvse daily from May 10 to June 13. The temperature, wind direction, and wind velocity were recorded hourly each day from 8 a. m. to 5 p. m., inclusive. 12 BULLETIN 273, U. S. DEPARTMENT OF AGRICULTUEE. Table IV. — Bates and numbers of newlii hatched caterpillars caught on screen, with tem- peratures, direction of wind, and velocity of wind at Plum Island, Mass., in 1914. Date and time removed. May 20, a. m. and p.m.. May 21, a. ra May 22, a. m May 23, a. m. and p. m.. May 24, a. m. and p.m.. May 25, a. m. and p. m.. May 26, p. m May 27, p. m May 28, p. m Total Number of caterpillars caught. Maximum tempera- ture. Minimum tempera- ture. Averacre between 9 a. m. and 5 p. m. Prevailing wind direc- tion. "Wind veloc- ity, in miles per hour. AV.andNW. NE. toSE., E. and SE.. AV , AV SAV SAV. and W AV. toS.... NAV 2 to 6 7 to 11 2 to 10 10 to 23 11 to 16 12 to 21 7 to 20 4 to 12 11 to 20 A total of 42 caterpillars were removed from this trap between May 20 and May 28 — a very short season in comparison to the records at Salisbury Beach in 1913. Five of these larvae Avere bloAvn on the screen by winds from the east and south, while the remaining 37 came with the winds from the Avest. Larvae were caught during the morning of May 20, when the velocity of the wind for that period ranged from 2 to 4 miles per hour, and on the follow- ing day, when the temperature ranged from 53° to 60° F. SCREEN EXPERIMENT ON ISLES OF SHOALS, N. H. The Isles of Shoals (PI. II), located 6 miles off the coast of New Hampshire and Elaine, was selected as an ideal place in Avhich to establish a trap for tests in long-distance spread. An elevated spot on Appledore Island Avas chosen and a screen erected in April, 1913. It Avas not known at that time that Appledore, comprising about 100 acres, was infested, but subsequent scouting in the summer shoAVod that this Avas the case. This island contains a varied groAvth of low shrubbery, some of which is very favorable food for gipsy- moth larvae, but the infestations were slight and covered small areas well scattered over the territory. The screen was examined at intervals during that year from May 1 to June 3 and 14 first-stage larvae were removed during that period. Some of these larvae were alive on removal and Avere taken on days when the Avind blcAV directly from the mainland, Avdiich indicated that many of them came from that source, hue there Avas room for question because the island on Avhich the screen Avas located in 1913 was later found infested. To eliminate any possibility of error in drawing conclusions another island in this group, namely, Lunging, was selected for the experi- ment in 1914. It is the most western of the isles and located so as to Bui. 273, U. 5. Dept. of Agriculture. Plate IV. Fig. 1.— Screen on Lunqinq Island, Isles of Shoals, N. H., on which 57 Larv/e WERE Caught During May, 1914. Ocean in Background. Vegetation in Foreground is Unfavorable Food for the Larv/e. (Original.) Fig. 2.— Screen on Holts Hill, Henniker, N. H., where Two Larv/e were Caught in 1914. (Original.) SCREEN TRAPS FOR WIND-BLOWN GIPSY-MOTH LARV/E IN NEW HAMPSHIRE. DISPERSION OF GIPSY-MOTH LAEV^ BY THE WIND. 13 receive the full force of the northwest, west, and southwest wmds directly from the mainland. This island is composed mostly of bare rock, with small patches of soil. There is little favorable vegetation for gipsy-moth larvae, and three careful examinations made in the spring of 1914 revealed no egg clusters. Four hundred and fifty square feet of wire screen was used, and an extra screen was made of cotton cloth tacked to a frame 40 feet long and .3 feet wide, and tanglefoot apphed. It was anchored at an incline on the rocks, facing the west. On this island there is one cottage which afforded quarters for Mr. Harry L. Blaisdell, who was in constant attendance at the screen (PI. IV, fig. 1). Arrangements were also made with the United States Life-Saving Service for transportation to and from the islands on their regular trips to Portsmouth, N. H. Mr. Blaisdell recorded the weather conditions hourly from 8 a. m. to 5 p. m., using a compass, thermometer, and hand anemometer. He also examined the screens continuously and made careful notes on the time and wind conditions when each caterpillar was caught, which made it possible to detemiine the source from which it came. Practically all the larvae found on the screens were alive, and in most cases they were removed very soon after entanglement. Table V. — Number of caterpillars caught at Lunging Island, Isles of Shoals, N. H., in 1914, icith direction ajid distance from xchich they came. Date. Other islands of the Isles of Shoals, E., NE., SE., and S., Jand g mile. Kjttery, Me., NW., 7 miles. Wallis Sands and Rye, N. h:, WNW., 6 miles. Rye Ledse and Little Boars Head, N. II., W., 7 miles. Seabrook, New Hamp- shire, and Ma=:sachu- setts line, WSW.,11 miles. Plum Island (north end), Mass., SW., 13i miles. 1914. May 20 6 6 1 1 May 21 May 22 May 23 8 20 3 May 24 1 " May 2J 2 1 1 1 9 May 26 May 31 June 1 I June 2 Total 25 1 28 3 1 9 Sixty-seven larvae were removed from the screen and cotton cloth between May .20 and June 2. The wind data given in the table were taken on the island, but comparison was also made with the records taken at Plum Island, Mass., 13^ miles to the southwest. It was thought that the air currents on the shore of the mamland where most of the small larvae were picked up would be much changed from their course before reaching the islands, but onl}" slight differences were noted by the two observers during the period. 14 BULLETIN 273, U. S, DEPARTMENT OF AGEICULTUEE. It will be noted, according to the records taken on Lunging Island, that nine larvae came from the southwest, that is, from Plum Island, 13^ miles away. The weather records at the latter place indicate that conditions were such tliat at least 18 of those caught on the screen at Lunging Island came from Plum Island, or more distant points in that dhection. Tliirteen and one-half miles is the maximum record for distance that larvae have been carried by the wind in all experiments thus far conducted by us, yet it is probable that the limit has not been reached. The locations for securing such records, together with suitable con- ditions with reference to infestations, are rare. DISPERSION EXPERIMENTS ON THE HILLS OF NEW HAMPSHIRE IN 1913 AND 1914. In the scouting operations conducted by Mr. L. H. Wortliley in the outside territory infested by the gipsy moth, numerous infestations were bemg found on hilltops. It could not be ascertained whether this spread was the residt of the wind, vehicles, or other possible car- riers — man or animals. These hills contained woodlands, orchards, pastures, mowings, and roads over which there was more or less travel, making spread possible artificially. Thinning was practiced in these infestations, egg clusters were creosoted, and bands of tanglefoot were placed on the trees to keep the first-stage cater- pillars hatching near the ground from ascending the trees, thereby decreasing their chances of spread by the wind. To investigate the source of these hiUtop infestations, one hill was selected in each of four towns in New Hampshire, namely, Andover, Henniker (PI. IV, fig. 2), Hillsboro, and Troy, and large screens erected thereon as traps for the small caterpillars. (See also Plate VII.) It seemed desirable to know if these colonies were the result of windspread, and whether caterpillars were being carried from hill- top to hiUtop, valley to hilltop, or vice versa, so that scouting and control methods could be changed accordingly. The screens were constructed of f-inch mesh poultry wire 3 feet wide. This was stapled to posts set in the form of a T. The posts were about 12 feet high and guyed with No. 13 wire. The wire was fastened in two widths, one above the other, making a screen surface 6 feet wide. The top selvage ranged from 10 to 12 feet from the ground owing to the level of the land. The screens were erected with one wing 35 feet long, facing the north and south, and another, 40 feet long, facing east and west. The total number of square feet exposed to the winds in each screen was 450. There were known infestations one-tliird to two miles in all directions from the screens in the foregoing towns, but negative results were DISPERSION OF GIPSY-MOTH LAEV.^ BY THE WIND. 15 secured on those at Hiilsboro and Troy, N. II., the infestations in the neighborhood of which were fewer and at longer distances. In 1912 several large gipsy-moth colonies were found in Andover and Henniker, N. H. On Beech Hill, Andover, about one-tlm^d of a mile from where the screen was erected, over 30,000 larvae were destroyed in the summer of 1912, This colony and surroundings were carefully treated, so that the infestation was nearly extermi- nated in 1913 and 1914. AU the smaller colonies found in the town also received thorough treatment. Similar conditions prevailed in Henniker. The screens were examined at periods during the dispersion season of 1913 and 1914. No positive results were secured during the first year, but one hving larva was removed from the Andover screen May 27, 1914, at 2 p. m. May 26, 1914, a dead larva was removed from the Henniker screen and a living specimen June 13, together with the molted sldn of a second-stage larva. It is difficult to state from which direction the first larva was blown, but the second came from the northwest or west as the wind blew from those directions during the two days previous. The 3 first-stage larvae caught on the Andover and Henniker screens in 1914 proved conclusively that the dispersion among the hiUs is brought about most extensively by the wind. Tliis is more emphatic when one considers the small air space covered by these traps and the small sources of infestation in these localities. ALTITUDE EXPERIMENT WITH AVIATION OF SMALL CATERPILLARS. At the close of the season's work in 1913, after securing some fair records on the distance that larva? are borne by the wind, it seemed advisable to get data on the possible altitude they reach during their transportation. After numerous inquiries and investigation a suit- able place for such an experiment was found in Merrimac, Mass. (PL VII). This consisted of the standpipe (PL V, fig. 1) for the stor- age and pressui'e of the town water. It is located on a small hill 240 feet above sea level, and permission to erect a small wire screen on the top of the tank was obtained from the selectmen of the town. The tank was 55 feet high from the level of the summit of the hill and a screen of f -inch mesh poultry wire was stapled to upright 2 by 4 posts 5 feet long. The top selvage of the wire (PL V, fig. 2) was 5 feet above the top of the tank, totaling 60 feet above the ground. The screen was 36 feet long and contained 144 square feet, while the tank was 40 feet in diameter, allowing ample space for the examiner to get around the ends. It was set up so that the sides faced the east and west. 16 BULLETIN" 273^ U. S. DEPARTMENT OF AGRICULTLTEE. Near the tank were a few low trees consisting of white pine and gray birch; these were cleaned of egg clusters and the trunks tangle- footed in order to prevent the larvae resulting from overlooked egg clusters getting into the tops of the trees. The country about this hill on all sides is generally infested (PI. VI), so that almost all winds tempered to 50° F. or above bore caterpillars. The weather data used for comparison with the daily catch here was taken at Plum Island, about 10 miles east. Table VI.- — Dates and numbers of caterpillars removed from screen at Meirimac, Mass., in 1914; also general direction of the loind during the period. Date and time removed. Number borne by west winds. Number borne by east and south winds. Direction of wind at time of examinations and a few hours previously. May 19, p. m 1 9 22 '% 8 32 2 (') 3 12 2 n 1 NW. and W., 8 a. m. to 5 p. ra. May20, p.m NW. and W., May 19, 8 a. m., to May 20, 2 p. m. May21 May 22, p. m 44 NE., E., and SE., May 20, 3 p. m., to May 22, 5 p. m. W., NW., and SW., May 23, 8 a. m., to May 27, May 23, p. m 10 a. m. Do. Maj' 25, p. m Do. May 26 May 27, p.m Do May 27, p. m Changed to S., May 27, 11 a. m., to 5 p. m. W. and NW., May 28, 8 a. m. to 5 p. m. May 29 Maj' 30, p. m SW. and W., May 30, 8 a. m., to June 1, 11 a. m. May 31 June 1, p.m S. and SE., June 1, 12 m. to 5 p. m. W. and NW., June 5, 8 a. m., to June 6, 5. p. m. Total 91 50 1 Some of these larvse were caught before thi; date, as 7 were dead on removal. 2 No examination. 3 Number accredited to different winds estimated owing to changes in direction. One hundred and forty-one larvfe were trapped on this screen of 144 square feet between May 19 and June 2, inclusive. A few of these undoubtedly were blown on before May 19, as seven of the nine were dead on removal, indicating that the aviation of the small larvae began probably two or more days earlier in Merrimac than along the seacoast. With apparently an equal infestation on all sides of this screen one-half mile or more distant, almost twice as many larvae were trapped as a result of prevailing winds from the west as from the east. The data secured in this experiment indicate that all wmds during the day in a section generally infested bear caterpillars excepting the north, which was not recorded during this period. Considering that practically one larva was caught per square foot 300 feet above sea level, it is only by conjecture that we can estimate the maximum height reached by them and the length of time they remain suspended under average conditions. It is a known fact that large numbers of Bui. 273, U. S. Dept. of Agriculture. Plate V. / ...4*0-^ -*-'^^ ••' ^ "■^^'^^fl'^ii ^^^^^^^^g^^M^^^i^^i^ii jg ..--'it. : •; ^r ^-7 - %i.t ' \fei. Fig. 1.— Showing Standpipe, Merrimac, Mass. Arrow Points to Screen Trap ON Top. (Original.) Fig. 2.— Screen on Standpipe, Merrimac, Mass., at Close Range, where 141 Larv^ were Caught During May and June, 1914. (Original.) SCREEN TRAP FOR WIND-BLOWN GIPSY-MOTH LARV/E, MERRIMAC, MASS. Bui. 273, U. S. Dept. of Agriculture. Plate VI DISPEESION OF GIPSY-MOTH LAEV^ BY THE WIND. 17 isolated infestations are scattered in hill and dale over 19,378 square miles of the New England States (PI. VII), excepting Vermont,* and a comparison of such information with the foregoing data at once suggests the source of practically all spread of this species. OBSERVATIONS ON WIND DISPERSION AT SALISBURY BEACH, MASS., IN 1913 AND 1914. Salisbury Beach, owing to its separation by extensive marshes from infested woodlands and to the hmited amount of favorable food plants for gipsy moths thereon, was selected as a desirable location to be cleaned up and watched for reinfestation. The south end of the beach (PI. II), IJ miles in length, was scouted and the egg clusters were creosoted in 1913 and 1914. Beach plum was the predominant shrub growing on the sands and was not fed upon by the caterpillars. Bayberr}^, wild rose, wiUow sprouts, balsam poplar, and other small growth thrived to a certain extent and are favored food plants. The scouting in 1913 residted in the finding of 144 egg clusters distributed over the beach in a few clumps of shrubbery. The con- trol work was followed up during the summer by frequent examina- tions for caterpillars, which resulted in their discovery in seven or more places other than where egg clusters were located. From one to six caterpillars were found in small isolated clumps of bay- berry and wild rose in no less than six spots near which there were no egg clusters. These patches of favorable food afforded good traps for the newly hatched caterpOIars, which were being carried from the infested woodlands 1 mile to the west, as did the tanglefooted screen on which such large numbers were caught the same year. Seventy-seven egg clusters were found and creosoted in 1914 in the same area as was scouted in 1913. Thus it will be noted that there were slightly more than one-half the number of egg clusters found in 1914 as in 1913, in spite of the rigid control measures practiced durmg the former year. The location of the beach with respect to the mfested woodland across the marshes was ideal for heavy reinfesta- tion each year, and this condition prevails in apple orchards, wood- lands, and shade trees that are cleaned within the gipsy moth infested area and become reinfested annually by means of the wind. OBSERVATIONS ON ISOLATED TREES AND YOUNG APPLE ORCHARDS FOR REINFESTATION. Twelve isolated trees in the midst of cultivated fields and mowings were selected in the early spring of 1914; these were cleaned of egg clusters, trunks tanglefooted, and later examined for reinfestation b}'' the wind. These trees were selected in towns about Merrimac, Mass., 1 During the winter of 1914-15 a few small infestations were discovered in Vermont. 18 BULLETIN 273^ U. S. DEPAETMENT OF AGRICULTUEE. where there was a general infestation. Some of thcni were later dis- carded owing to the discovery of overlooked egg clusters that had hatched. On the trees retamed there were from 2 to 8 and in one case 11 large caterpillars above and below the tanglefoot, indicating that there was reinfestation by the wind. The top of a medium to large sized apple tree would offer more resistance to the wind than did the screen on the Merrunac standpipe, upon which was lodged approximately one first-stage caterpillar to every square foot of wire exposed. Some further observations confirming these conclusions were made by Mr. C. E. Hood in July, 1914, in two young apple orchards in Merrimac (PL VI) and West Newbury, Mass. These orchards were 2 and 4 years old, respectively, and clean cultivation was practiced m them, A large number of the trees were examined in the center of these orchards, which were in some cases 500 yards from the nearest mfestations on larger trees. Thirty-four large larvag were found in selected areas of the two orchards, not more than one of which appeared on a single tree. No egg clusters could be found, and it was concluded that the infestation was due to windspread. Mr. J. V. Schaffner, jr., has also reported a similar incident from Dover, Mass. SMALL LARV^ BLOWN INTO THE OCEAN ANNUALLY FROM INFESTED WOODLANDS NEAR THE COAST. At the rate of 266 first-stage larvse caught on 450 square feet of tanglefooted wire at Salisbury Beach in 1913 (being located 1 mile from the nearest infested woodland), a basis is given for estimating the huge numbers borne into the ocean each year. At the same ratio a continuous screen 1 mile long and 6 feet high would have caught 18,726 larvae. Judging from the altitude, 300 feet above sea level, where quantities were caught at Merrimac, Mass., it is apparent that this number should be fifty times as much, or 936,300 per mile. There is now upwards of 450 miles of coast line infested in New England (PL VII), and figuring that general spread by wind occurs over one-half of this distance, the ocean swallows up annually over 210,000,000 small larvse. The number is probably much greater than is here indicated, as it is not known how many of the small larvae passed through the wire meshes of the screen which is used as a basis for these figures. These calculations, which are partly theoretical, serve to indicate what might have been the result with regard to the area infested had the moth first been introduced 100 miles farther westward than Med- ford, Mass. It also emphasizes the necessity of keeping this moth confined to New England territory, if rapid and general spread over the United States is to be prevented. DISPERSION OF GIPSY-MOTH LARV^ BY THE WIND. 19 FEEDING OF LARV^ PRIOR TO DISPERSION BY THE WIND. Several of the first-stage caterpillars of various sizes caught on the screens during 1913 and 1914 were selected and examined for the presence of plant cells in the alimentary tract. As aU caterpillars blown any reasonable distance by the wind are newly hatched, first- stage specimens, it is impossible to ascertain from a superficial ex- amination which have taken food and which have not. Before the examinations were made it was necessary to fix the material, stain with eosin and methylene blue, section with a microtome, and mount on slides, and this work was performed at Bussey Institute, Harvard University, under the direction of Dr. II . W. Glaser of the Bureau of Entomology. Caterpillars were selected diu'ing the two seasons which had been borne by the wind a distance of 1, 2, and 6 or more miles, respec- tively, from the nearest source of infestation. Results were as follows : Table VII. — Percentages of small caterpillars having Jed which were caught on screens during 1913 and 1914- Year. Locality and distance carried by wind. Number of caterpillars that had fed. Number of caterpillars that had not fed. 1913 Salisbury Beach, Mass., 1 mile from infested woodland 13 1 2 7 7 1914 Isles of Shoals, N. H. Infestation in brushland aroimd screen Plum Island, Mass.; 2 miles from infested woodland. ... 5 Isles of Shoals , N . H .; 6 miles or more from infestation on the main- land to the westward 8 Merrimac, Mass.; J to 1 mile from heavy infestation in all directions. . Total 2 23 22 Of the 45 small caterpillars prepared and exammed during the two years that experiments were conducted, 23, or 51 percent, had con- sumed a very small amount of food, while the remainder, 40 per cent, -showed no signs of it. COMPARISON OF WEATHER DATA BETWEEN PROVIDENCE, R. I., AND AMHERST, MASS., WITH REFERENCE TO DISPERSION IN 1913. The period m which first-stage larvae were borne by the wind in 1913 in eastern Massachusetts was from May 9 to June 5, inclusive. After comparing the hourly wind direction from 7 a. m. to 6 p. m. when there was no ram and the tem|3*>rature was 50° F. or above at Providence, R. I./ and Amherst, Mass.,^ one notes a slight difference in the total wind movement in the various directions. For mstance, in Providence the wind blew from the northwest 92 hours, from the 1 The weather records at Providence were secured from the oflBce of the U. S. Weather Bureau, and those at Amherst from the Experiment Station. 20 BULLETIN 273, U. S. DEPAETMENT OF AGKICULTUKE. aortli 17, from the northeast 22, from the east 3, from the southeast 51, from the south 37, from the southwest 1, and from the west 9 hours duruig the foregomg period. In Amherst, Mass., it blew for the same period from the northwest 50 hours, from the north 51, from the northeast 9, from the east 2, from the southeast 30, from the south 31, from the southwest 3, and from the west 46 hours. Providence was first found mfested by the gipsy moth m 1901, and during the scouting season of 1913-14 egg clusters were found as far west as Woodstock, Pomfret, and Brooklyn, Conn. — a distance of 36 miles, or an average spread of 3 miles each year. It is apparent that this general infestation was not accomplished by direct east wmds, as there are few recorded each year at the proper period, but by a combmation of northeast and southeast winds transporting the larvae m a southwesterly and northwesterly direction from 1 to 10 miles, thus gaining an average of 3 miles each year directly west. Although this insect has not yet reached Amherst, Mass., in its general sweep across the country, it occurs withm a few miles to the eastward. A perusal of the wmd records for the dispersion period of 1913 shows tlu'ee times as many hours of north winds and about one-half as many northeast and southeast winds combined as are recorded m Providence, R. I. The total movement of northwest and west wmds, which blow the larvae back into the infested terri- tory each year, is practically the same in the two locahties. From this data it appears that the increase m the amount of north wmds and the decrease in combmed northeast and southeast winds at Amherst, when compared with the wind records at Providence, may result in a somewhat more rapid southwestern advance of the insect if it becomes established in the Connecticut Valley. More rapid southern and western spread is likely if the infestation reaches the Central States! SPREAD OF THE GIPSY MOTH IN NEW ENGLAND. The gipsy moth was introduced at Medford, Mass. (PI. VII), m 1869, from which pomt it has spread gradually over large areas m Massa- chusetts and other New England States, excepting Vermont.^ Small mfestations have also been located in New York, New Jersey, and Ohio, but these colonies are fast being exterminated. Extensive efforts to prevent spread have been carried on in Massachusetts by the State since 1890, with suspension of appropriations and efforts only from 1900 to 1905. Other mfested States have appropriated smaller amounts which have aided in the suppression work. Con- gress began making appropriations in 1906, and these aimual sums have been expended to prevent spread and help control of the gipsy moth and the brown-tail moth. ' See, however, footnote on p. 17. DISPERSION OF THE GIPSY MOTH NEW ENGLAND AND AREA QUARANTINED - 1914. DISPEKSION OF GIPSY-MOTH LAEV^ BY THE WIISTD. 21 Table VIII is here presented to show the annual advance of the moth before and after control was attempted. This indicates that the winds are chiefly responsible for the present conditions. Table VII T. — Spread of the gipsy moth in various directions htj series of years from five years after its introduction at Medford, Mass., to 1914. Periods of years. NE. N. NW. W. S\V. S. SE. Average spread per year by periods. 1875-1890 Miles. 15 5 27 41 114 Miles. 10 4 11 (iO 16 Miles. 5 5 12 25 28 Miles. 10 4 10 19 15 Miles. G 4 12 15 24 Miles. 4 5 14 16 26 Miles. 3 4 62 11 Miles per year. 0.5 1891-1899 1900-190S 3.5 1906-1909 6.7 1910-1914 6.4 Total 202 101 75 58 61 65 80 2.3 Average spread per year in 5 2.5 1.9 1.5 1.5 1.6 2 It will be noted that the spread from the original center was very slow, averaging only 0.5 of a mile per year up to the beginning of the campaign in 1890. It was then held in control for a period of about nine years, during which time there was very little advance into new territory. The conditions in the old infested area were also much improved. This shows the necessity for keeping the moth under complete control if windspread is to be prevented. During the period from 1900 to 1905, when no work was being done, the moths spread at the rate of 3.5 miles per year. This had increased to 6.7 miles per year from 1906 to 1909, but this ratio has decreased to some extent in the last period up to 1914, inclusive. The greatest distance gained in various directions has been to the northeast, an average of 5 miles per year for 40 years, which is the result of the favorable southwest winds. The combination of south- east and southwest winds during the dispersion period has carried the larvae northward at the rate of 2.5 miles per year. The spread to the west, southwest, and south was accomplished at a slower rate, owing to the less favorable winds blowing; in those directions. SUMMARY. In 1913, as a result of the several experiments conducted by using tanglefooted screens and cloth for traps, there were caught on 977 square feet 289 first-stage larvae which had been borne by the wind one-eighth to 1 mile or more. In 1914 there were removed from 1,614 square feet of sticky surface 346 larvfe which had been blown from one-eighth to 13^ miles or more, as verified by the wind records taken at or near those points. Three larvae were also taken from two large screens on the hills in New Hampshire during 1914. 22 BULLETIN 273, V. S. DEPARTMENT OF AGEICULTURE. Considering the great numbers of lai'vse taken in these experiments, there can be no doubt that the wmd is ahnost wholly responsible for the general spread of this insect in New England, notwithstanding the fact that many of the former pubhcations teem with explanations of possible accidental or artificial spread by man and animals. The recent establishment by the Federal Horticultural Board of a quaran- tine on lumber products, Clu-istmas trees, nursery stock, and stone moving from the infested territory, until inspection has been made, has greatly relieved the danger of a general distribution of the gipsy and brown-tail moths over the United States, but the sources of dis- persion of the gipsy moth by the wmd can only bo lessened by the effectiveness of the parasites and predaceous enemies, together with hand methods of control. To prevent continual spread by the wmd into new territory the badly infested areas near the border must be brought mider control either by natural enemies, or hand methods, or both. Natural enemies, however, are now playing an important role in the control of this insect in the greater area of the inside infested territory. The larvae are sufficiently active and allow themselves to be trans- ported by the wmd at temperatures of 55° F. and above, and have been caught at wind velocities varymg from 2 to 23 miles per hour, although more active spread takes place when the temperature ranges'from 65° to 85° F. and when the velocity reaches 8 miles or more per hour. Larvte are removed from their support and carried by sudden gusts of wind, whether they spin or not, when the tempera- ture reaches 50° to 55° F., at which temperatures they often start crawling. The records also show that larvae have been caught at times when winds were blowmg from all du-ections except the north— only a very few coming from the east, but the location of the screens along and near the coast materiaUy affected this condition. By far the larger numbei-s were borne by combuiations of the west winds as indicated on the screen at Merrimac, Mass., wliich was surrounded by a general infestation. The general progress of the species since its estabhshment at Medford, Mass., at the rate of 5 miles per year to the northeast and at the rate of 3 miles per year westward from Providence, R. I., since its first appearance there in 1901 tends to verify the data that have been collected in connection with the screen experiments. BIBLIOGRAPHY. 1893 Wachtl und Kornauth. Beltrage zur Kenntniss der Morphologie, Biologie und Pathologie der Nonne (Psilura monacha). In Mittheilungen aua dem forstlichen Versuchsvesen Osterreichs, v. 16, 38 p., 3 pi., 8 fig. 1894. Cholodkovsky, N. Ueber die sogenannten "Aerophore" der Nonnenraupe. In Forst-naturw, v. 3, no. 5, p. 240-243, 1 fig., May. DISPERSION OF GIPSY-MOTH LAEV^ BY THE WIND. 23 1896. FoRBusH, E. H., and Fernald, C. H. The Gipsy Moth. (Rpt. Mass. Bd. Agr.), 495+h'iii p., 66 pL, 37 fig. 1897. Ingenitzky, I. Zur Kenntniss der Driisenhaare der Noimenraupe {Ocncria Monacha L.). In Horae Soc. Ent. Ross., v. 30, p. 129-134, 1895-'96. 1912. EscHERicH, K. Die Bedeutung der "aerostatischen " Haare der Spiegelraupe. In Naturwissenschaftliche Zeitschrift fiir Forst- und Landwiitschaft, v. 10, no. 2, p. 82, February. 1913. Burgess, A. F. The Dispersion of the Gipsy Moth. U. S. Dept. Agr., Bur. Ent., Bui. 119, 62 p., 16 pi., 6 fig. 1913. Riley, Wm. A. The so-called aerostatic haii-s of certain lepidopterous larvae. In Science, n. s., v. 37, no. 958, p. 715-716, May. 1914. Shcherbakoa\ T. C. Observations on the gipsy moth (Lymantria dispar L.) Translation from the Russian. In Proc. Mus. Nat. Hist. Taurican State Zemstvo, a-. 3, 1914. WASHINGTON : GOVEBNMENT PRINTING OFFICE : 1915 Syracuse, N. Y. PAT. IAN. 21, 1 90S Mill i!S . iiiii ^iiii m DDDDflflflflb^a