«^ 4 1"' LI. 5 -^i9 • tasSSI. MB ^ra^WTT""^ I -^ tt- Tp T*--^- *i - ^^i ^L WtL ^r v«B[; C F < .,„,, -«i, ■" K r c, a .L *^^ — t::=3t> i^S flVK^ ^€ r-^^^f i^^^Bg jr'w'^ ra 2 wS^^^^S^ •rrl.; 1 rr< c« « r «t^i < i^mm- m< (/ 4^\ S^ 1= -^^^U^^^^^^^^^Ll ^K| ._^_, . r~» S- ^l~Ml Y? y — - ^R' ft iiT ^^'"-" r If" it — ^^r—" Sr^ ■■ H- ; 1 P- '^^ " f ■^■^ ""V' - ""la" if— < i-_^H \ C ■■i: <■■( {*" P-r , X" F ■^- * ' p- - -. ^f^ J^P? lT — ■» =^am T t k VV _ t J_ N-^ N^ .w.-^- « ^Tr " ^ K k' *l l» '\ ,,,, 'h p. ^ V -W »f~ zj ■ r sifir ^^=P ip4 sc«an« — - — f~ 3rT^* r-fc "2*^ — *^ { # ■ ( ft * 1 r ^— '%^ ^ ' »: — ^■iiL-' •■'■ <'-'r ^^^T Vi^Fl ^K^'CKi ■ r.iPf^^Si: V fe »L.X ■K'JS-^ ■ P~^ ^B^ v% — ^^r~ p^^ W ,r- 1 T%' w 5-^ Z//?^ UNITED STATES DEPARTMENT OF AGRICULTURE S)J9^^^L % BULLETIN No. 804 Contribution from the Bureau of Entomology L. O. HOWARD, Chief JTU^'^'^-Zt. Washington, D. C. PROFESSIONAL PAPER March 16, 1920 A STUDY OF THE BEHAVIOR OF BEES IN COLONIES AFFECTED BY EUROPEAN FOUL- BROOD^ By Arnold P. Sturtevant Specialist in the Bacteriology of Bee Diseases CONTENTS Page Introduction 1 Procedure 5 Observations 8 Summary of previous experiments-- 15 Supplementary observations 17 Study of naturally infected colonies 17 Behavior of bees in cleaning contaminated cells 18 Possible infection through queen 19 Page Supplementary observations — Contd. Distribution of introduced in- fected material 20 Age at which larvae are in- fected 21 Microscopical bacteriological observa- tions 24 Summary and conclusions 26 Literature cited 28 INTRODUCTION The brood diseases of bees cause annually large losses of bees and consequently of the honey croj). The predominant attitude among beekeepers has long been how best to eradicate an invading bee dis- ease after the attack has been made. They depend upon this pro- cedure, because little is known with any degree of certainty concerning the natural conditions which might prevent or control the onslaught of the disease. As a result of this attitude, much more importance has been placed on the significance of apiary inspection and police- power laws and of purely remedial treatment, the reasons for which in many cases are imperfectly understood. But the old adage " an ounce of prevention is worth a pound of cure " has yet to be refuted, jjarticularly with regard to beekeeping. In the realm of human ^A series of investigations was started in the spring and summer of 1918 by the Oflace of Bee-Culture Investigations, Bureau of Entomology, for the purpose of making an intensive study of European foulbrood of bees, primarily from the standpoint of insect behavior in relation to the disease, correlated with the facts and practical observations already Imown to the beekeeper. This paper, which was submitted for publication January 13, 1919, is a preliminary report on the beginning of the in- vestigation. 134440°— Bull. 804— 20— —1 2 BULLETIN 804, U. S. DEPAETMENT OF AGRICULTURE medicine, for the last two decades at least, this precept has been gaining strength so that to-day preventive medicine stands on a par Avith, if not above, most of the other branches of medicine. Why is it not logical to apply this principle to the control of bee diseases ? Ever since European foulbrood of bees was first recognized (in 1894), in New York State, as a distinct brood disease, there have been much controversy and speculation concerning the etiology of the disease, the means of transmission, the method of spread, and, result- ing therefrom, the question of control. From the laboratory stand- point, the etiology of the disease has been worked out quite definitely bacteriologically (12)\ But as yet Bacillus 'plvA.on^ the accepted cause of European foulbrood, never has been gi'own in pure culture on artificial media, although it has been definitely identified as the cause of the disease. This precludes any further advance along this line of attack for the time being. From the side of practical experience, there have been recorded large numbers of observations, many of them of a similar nature. These observations have led to many accepted practices, as, for in- stance, the use of Italian bees and strong colonies in combating the disease. Although the weight of numbers tends to give substantia- tion to observations, the scientific explanation of how these things are true never has been studied carefully and coordinated with the practical side into an epidemiological study of the colony under dis- ease conditions in European foulbrood. \ The history of bee diseases has developed mainly along two lines. The scientific side has been concerned principally with determining the causes of the various diseases microbiologically, the method of diagnosis, and conclusively differentiating them. These facts have been described sufficiently in various bulletins of the Bureau of En- tomology and will not be discussed here. From the practical side, countless observations have been recorded, largely in the bee journals, in which various manifestations of the disease and experiences with methods of treatment have been discussed. But in all this literature, particularly with regard to European foulbrood, there are few ob- servations on the disease and on the behavior of the bees in relation to it beyond simple description of symptoms. Early in the experience with European foulbrood it was learned by careful observers that strong colonies are essential in successfully combating the disease. Later the value of Italian bees was dis- covered. West (11), a NeAv York State apiary inspector, in giving what is one of the best early descriptions of European foulbrood, makes some pertinent observations on the disease. He states that when diseased brood is placed above a strong, healthy colony, with a queen excluder between, so that any healthy brood may emerge, 1 Reference is made by number in parenthesis to " Literature cited," p. 28. BEES IX COLONIES AFFECTED BY EUROPEAlSr FOULBROOD 3 the diseased larvae are cleaned out as this is taking place. The union with a healthy colony and the strength gained by the emergence of so many young bees gives the colony the stimulus to eliminate the disease. He notes, as have many other beekeepers since, that in August, when the buckwheat honey flow begins, the stronger of the diseased colonies are stimulated to clean up. Alexander (1) published a method of treatment for European foul- brood, the principle of which, after many varying failures and suc- cesses, is now the basis for the present method of treatment most used; that is, requeening with Italian stock. Alexander mentions the need of three factors: First, the necessity of requeening with young yellow Italians, as hybrids of Italian and black bees are prone to contract the disease in the first place and also are more likely to succumb to it; second, particularly emphasized, a period (at least 27 days, according to Alexander) of queenlessness in which to allow the bees properly to clean up the cells and polish them, preparatory for eggs of a new queen ; third, a factor which is mentioned only casually but which is equally important with the other two, the direction to unite tod strengthen diseased colonies before treating. So little em- phasis was placed on this that the majority of beekeepers overlooked it in using Alexander's treatment and therefore condemned the treat- ment as unsuccessful except in rare cases. In an editorial (8) in the same issue of the journal m which Mr. Alexander was writing, the question was raised as to why the period of broodlessness caused by winter, which is much longer than 27 days, does not always prevent a recurrence of the disease. Mr. Alex- ander answered this question by explaining that when the queen stops laying in the fall, the bees do not polish up the cells as they do earlier in. the season, and that some of the dried-down material may remain until the next spring. The opinion also is given in this editorial that Italians are more able to resist the disease than hybrids because they do more thorough work in house cleaning and are less inclined to rob. Phillips (6) makes the statement that "European foulbrood is more destructive during the spring and early sunmier than at other times, often entirely disappearing during the late summer and early autumn, or during a heavy honey flow," but gives no indi- cation as to how this takes place. The same year Miller (2) pub- lished his theory of the relation of the nurse bees to the spread of European foulbrood. He believes that the nurse bees suck up the juices of a freshh' diseased larva which has not become offensive, and then transmit the disease when feeding the healthy larvae. On this supposition he believes that if egg laying ceases for 5 or 6 days (" the period the larvae remain unsealed in their cells ") there will no longer be larvae in the proper condition for nurse bees to feed upon, 4 BULLETIN 804, V. S. DEPARTMENT OF AGRICULTURE nor healthy unsealed larvae to receive the infection, and the disease will thereby come to an end. Dr. Miller has been using a 10-day period of queenlessness in his treatment of European foulbrood since his accidental discovery that 10 days .were sufficient, but in a later article (3) in enlarging upon his nurse-bee theory he assumes that the larva is fed during a period of 5 days but is not effective as a carrier of infection during the whole time as probably no larva? are torn open until they are 2 or 3 days old, thus making it possible to shorten the queenless period even more. He admits that not all the dead, partially dried larvae will be cleaned out, but believes that it is only the fresh yellow ones which are infectious. He also states that nurse bees are not inclined to travel far on the combs, a fact which may explain why the disease may be found con- fined to one comb for several days before spreading farther. Dr. Miller seems to have overlooked several important factors which will be discussed later. Quite an extensive piece of investigation was carried on during the summers of 1915 and 1916 by the author at the Massachusetts Agri- cultural Experiment Station upon the effect of requeening diseased colonies with various strains of Italian bees. At that time the im- portance of strong colonies with the requeening had not been em- phasized so strongly and less attention was paid to that factor. The records show, however, that in a total of 50 colonies observed, cov- ering two seasons, of 10 strong colonies only 2 showed recurrence, while 1 was doubtful ; of 20 medium-strength colonies, 10 showed re- currence with 2 doubtful; of 14 weak colonies, 8 showed recurrence. In all these cases the new queen was not introduced until the colony was nearly or entirely clean. In the case of several of the weaker colonies it was necessary to strengthen them before requeening was possible, in order to save the colony. One or two of these, which were united and requeened with Italian stock, were the best colonies the next spring. Adding some strength to at least part of Miller's theory is a state- ment in a letter by G. C. Matthews, formerly of this bureau, who wrote in February, 1918, concerning his observations in California in 1914. He found that where the hives stood in rows of pairs the disease continued to spread down each row to corresponding members of each pair. This ceased when he rearranged his apiary so that the rows of hives were at least 10 feet apart, and alternate pairs of hives were turned at right angles. No pair was allowed to remain close to another facing the same way. This prevented the drifting of nurse bees, which he believes to be the method of spreading the disease. Furthermore, he found by introducing one Italian queen into the middle colony of an isolated row of hybrid bees that there was considerable drifting of nurse bees. Seven days after the brood BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD 5 from the Italian queen began to emerge, yellow bees were found on either side in several of the hybrid colonies. Speaking of uniting weak diseased colonies and requeening, Matthews writes: After two or three were put together, each stack of brood was given an Italian cell. When young queens commenced to lay there was still disease in many of those hives, but as the queens increased in laying the bees cleaned out an ever-increasing sphere of comb for a brood nest until they had the hives free of disease. But in no case, however long a hive might be queenless, did I see the disease cleaned out before a virgin appeared in the hive. In other words, a virgin had to be present before the bees would commence their job of cleaning up. Therefore, I see little to commend the practice of keeping diseased colonies queenless 21 days. A new bulletin by Phillips (7) has been issued recently by the Department of Agriculture. The fundamental idea emphasized is that " in keeping European f oulbrood under control it is far more important to prevent the disease from getting a foothold in a colony than it is to eradicate the disease afterward," This bulletin, aside from discussing symptoms and methods of treatment, states concisely for the first time the facts observed in apiary practice on which successful treatment is based, and without an understanding of which it is difficult for a beekeeper to use preventive measures with any success. The analysis of these factors of response in behavior to treatment, as stated by Phillips, has been used to some extent as a foundation for the present work on the behavior of the colony in relation to disease, in an endeavor to substantiate, with data obtained under controlled conditions, these facts that are constantly observed in apiary practice and, if possible, to eliminate confusion in methods of treatment. PROCEDURE Shortly after the middle of May, 1918, experiments were started in Ithaca, N. Y., at the Cornell Agricultural College. Through the kindness of Prof. J. G. Needham, head of the department of ento- mology, and others associated with him, the use of a small, isolated yard of bees and also of laboratory facilities was offered for the purpose of carrying on these investigations. This small apiary had been used previously in fruit-pollination studies and had no record of disease. The yard was admirably located in a naturally well- protected hollow beyond the college fruit orchards, about a mile and a half from the main college apiary or other apiaries, with high ground and woods intervening. The author and the Office of Bee- Culture Investigations are under deep obligations to the Cornell authorities for the assistance so cordially extended. Being in the buckwheat district, the general locality was well adapted to the work because of the desire for as late a main honey 6 BULLETIN 804, U. S. DEPARTMENT OF AGRICULTURE flow as possible in order not to have the influence of a heavy' honey flow until other factors had been studied. At Ithaca the main honey flow is generally from buckwheat, coming from the 1st to the mid- dle of August. Rather unfortunately for the best results from the experiments, however, the summer of 1918 was unusual in this sec- tion, for the abnormally heavy honey flow from clover necessitated finishing the work earlier than had been planned, owing to the great difficulty of artificially infecting colonies during the heavy honey flow. There were seven colonies in the original experimental apiary. At first it was intended to work on a larger scale, but the trend of the observations soon led to the plan of worlring in more detail and on a smaller scale. These colonies were moved some distance apart to prevent drifting and robbing. Some were divided and some were strengthened in an efl'ort to make a series of experiments on colonies of different strengths. The colonies were designated by letter and the combs of each colony by number. From time to time some of these colonies were artificially infected with diseased European foul- brood larvae from samples sent to the laboratory for diagnosis. Similar colonies were held intact and uninfected for controls. The infection was made by feeding diseased larvre macerated in sugar solution (about 50 per cent). For the preliminary experiments 10 larvae were fed in about 250 c. c. of sirup. Later, after the heavy honey flow had begun, it was necessary greatly to increase this dose in order to start the infection. The infected sirup was fed to the bees in sterilized glass petri dishes, placed on top of the frames and pro- tected by an empty comb-honey super placed on the regiilar hive- body with the cover on top. At the time of inoculation, the condition of each colony was noted as to age, race, condition and appearance of the queen, proportion of nurse bees to old field bees, the number of frames of brood with the amount in each, its age, sealed or unsealed; in other words, the condition of the colony with regard to factors known to be signifi- cant in resisting disease. In two colonies the infected sirup was slightly colored with harmless eosin dye to determine where the fresh sirup was placed and its ultimate disposition. At first daily obser- vations were made to determine the earliest appearance of disease, the period of incubation, the symptoms exhibited, and the rate of increase. By holding up eacli comb in bright sunlight so that the light shone directly on the larvre, it was easy to detect the first symptoms of the disease. All the healthy larvro had the characteristic firm, well- rounded, pearly-white, glistening appearance. The first effect of the disease, besides an abnormal uneasy movement, was a loss of the BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD i glistening character and a slight tinge of grayish or creamy dis- coloration which would not be noticed except in direct sunlight. These larvae showed only Bacillus pluton present when examined microscopically, as will be mentioned later. Soon after these first symptoms, however, the more noticeable symptoms appeared, such as a larva with its back out, the increase of the light grayish yellow color, and, later, the moist, melting appearance. A statistical record was kept of the number of larvae showing new disease at each observation, the number previously diseased that had been cleaned out in the interval since the previous observation, and those remaining over in the cells uncleaned for more than one period between observations. At various times observations were made of the behavior and types of bees engaged in cleaning up and the fate of the material removed. Great care was necessary in these obser- vations to disturb the colony as little as possible. On good days it was sometimes possible to remove a comb carefully from the hive and to watch the bees continuing at their work, and even to watch the queen laying eggs. An eight-frame observation hive containing a strong healthy colony was given a diseased comb from time to time and the bees were observed as they worked on it. One of the difficulties of the work was to find a satisfactory method of recording the desired data for each comb. At first the diseased cells were marked on the comb by a circle of red celloidin around the entrance of the cell. Although this dried rapidly, it proved unsatis- factory, as the bees, in their attempt to remove the foreign material, seemed to. remove both diseased and healthy larvae indiscriminately. Next small pins were used, inserted in the cell above the one showing disease. In this case the bees tore down the surrounding cells and completely removed the pins, many of which were found on the bottom board. Finally a method of plotting the diseased cells in a comb was adopted. An empty frame was laid off in inch squares by means of heavy black thread. This, used as a templet superimposed on a comb, aided in the location of the diseased and cleaned out cells, so that they could be recorded on a correspondingly ruled card (fig. 1) . Placing this over the comb, it was easy to locate exactly each cell and to determine how Jong the diseased material remained, thus aiding in following the course of the disease throughout its various stages. The only difficulty with this method was the tediousness of the obser- vations. Therefore, after the disease had become definitely estab- lished, daily observations of each colony were considered unneces- sary. Longer periods showed just as well what was happening in the colony. Also, after the disease had developed enough so that it could be definitely predicted whether the colony would recover or gradually be exterminated; observations of behavior under treatment were 8 BULLETIN 804, V. S. DEPARTMENT OF AGRICULTURE started, the method and degree of house cleaning being watched after the colony had been dequeened, strengthened, and requeened with good Italian stock. Note was also made of any recurrence of disease and // /^ /9 /^ /:5 /^ /y'/S Fig. 1. — Method of plotting the location and history of diseased bee larvae in the combs. O Freshly diseased larvae. O Cells that have been cleaned out. 9-Cells that have been cleaned out and filled with nectar. (J) Larva; remaining in the cells more than one observation period. The area of sealed brood was the amount present at the time of infection of the colony. apparent reason therefor. In other words, a complete study was made of the cycle of the disease and of the activities of the bees during its course. OBSERVATIONS COLONY G Race. — ^Hybrid. Queen. — 1917, dark and poor. Bees. — Workers and drones very dark, almost black, very excitable. Condition of colony at time of infection. — Brood iu foiir frames, a little less than half sealed, besides two frames of eggs. Bees covering about eight frames, medium strength. Slightly more field bees than nurse bees, because of having divided this colony, old bees returning from the division. Date of first infection. — May 28, 1918. Material used. — Ten diseased larvae from sample No. 5863, macerated in 250 c. c. of a 50 per cent sugar sirup. First appearance of disease noted. — May 31, 1918, three day3 after inocu- lation. Age of larvw first attacked. — Three to four days after hatching from the eggs. Colony G (fig. 2), hybrids, soon succumbed to the infection, the first diseased larva appearing three days after infection, the gross diagnosis being confirmed by the finding of Bacillus pluton on microscopic ex- amination. The spread of the disease was rapid, the disease being present hi only one comb on the third day and in seven combs on the seventh day. All of this early spread took place in brood unsealed BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD 9 at the time of infection. The first high peak of the disease coming on the nineteenth day was followed by a slight improvement, when for a time the house cleaning exceeded the occurrence of fresh dis- ease. This was probably due to the stimulus of the increasing honey flow. But as soon as ihe next series of eggs hatched, the disease" again gained the upper hand, reaching another higher peak on the thirty-first day, at which time it was deemed necessary to start treat-' ment. It had become evident that the colony was being overrun by' the disease. More and more dead larva3 were being allowed to re-^ 134440°— Bull. 804—20 2 10 BULLETIN" 804, V. S. DEPARTMENT OF AGRICULTURE main in the cells for several days without being cleaned out. Also more larvae nearly ready for pupation were being affected. Most of these instead of remaining coiled were inclined to extend on the lower side wall in a brownish gray, slimy mass and exhibited a ten- dency to be viscid. At this stage of decomposition, when a stick is inserted the mass forms a coarse granular band for a short distance and then breaks so as to form droplike masses, but does not stretch out in a fine thread. These larval masses dried down to rubbery dark brown scales something like American foulbrood scales in ap- pearance, but different in consistency. These scales could be removed quite easily and would bend like a piece of partially granular old rubber. They also lay irregularly placed in the cells, often spirally extended, while American foulbrood scales are uniformly on the lower side wall. The bacteriological explanation for this abnor- mal characteristic will be discussed later under bacteriological observations. The predominance of these rubbery masses and scales increased as the disease progressed and the bees seemed to make little attempt to clean them out, even after the queen was caged on the thirty-first day, thus shutting off any increase of fresh larvss, or even after the queen and all queen cells were removed on the thirty-seventh day. On the thirty -ninth and also on the forty-first day, five and four frames, respectively, of emerging brood and Italian bees were united with this colony, but it was not until a new Italian queen, confined in a cage, had been hung in on the forty-fifth day that a final com- plete cleaning up was made. This new queen was not accepted, however, and a young queen was raised from the brood that was added to this colony, so that fur- ther obser\"ations were ended here although the virgin queen was killed and another Italian queen introduced. This colony was re- ported healthy, however, about the middle of August. The hybrid bees seemed to lack ambition to fight the disease. When combs were removed from the colony, the bees never were ob- served to be working in the cells, and paid little attention to ma- terial partially drawn from the cells and crushed. COLONY r Race. — Italian, possibly with some slight hybrid blood. Queen. — 1917, fairly good condition. Bees. — Workers, good color, fairly quiet, drones inclined to be darker. Condition of colony at time of infection. — Brood in three frames, a little more than one-third sealed. Bees covering about six frames. Build- . ing up well. Proportion of tield bees to nurse bees about equal. Date of first infection. — May 31, 1918. Material used. — Ten diseased larvfe from sample No. 5874, macerated in 250 c. c. of a 50 per cent sugar sirup. First appearance of disease noted. — June 4, 1918, four days after in- fection. Age of larvw first attacked. — Four days after hatcliing from the egg. BEES IN" COLONIES AFFECTED BY EUROPEAlSr FOULBROOD 11 Colony F (fig. 3), which was the next one to be infected, although not as strong as colony G, was of Italian stock and did not show the api^earance of disease until one day later. On the fourth day one cell appeared in each of two combs. It was not until the twenty- t^ fti f<; (^ =^ ^ =? _, ^^' "' ^'' / ^^ •"' / ,^-'' ' / ,^- -^ ,,'' / __^ ^'' \ . / ^^- '' XI / / ■;>' , . -^ , -^ / u ,.--''' ^N y \ ' ^^ ^ s ^.^ \ y \ N y N \ y y \ y^ ^ ^^ \ ^ -^^ \ s ^ \, \ ~~^ \ ^ ~-- ^ \ "^ ^ \ ^ ^~^ ^x '^^ \~^ \ 1^ k -•^>^ \ ^ ^^, ^ "■^ \ J A ^^^; Vi- ki \ \ / k V *> ^d \ \ \ V ^ / ^^f ^ fr^( ^^^ ^^ ■^ \ N \ ^ ,^^^ (n^^ 'ffi f^ ^ \^ ^ ^, \ ^^ 1^^ Is? ^1' 4 ^$x; K ^ ■^ -^x N 1 N h'5^ ^^s ^ ' ^^ H-f^ N^ "x N ^^ ^. ^J^ VllS .5?< ^' ii^' i ^^ ^^.^ ?}? ^i'i i^l^ ??^ \v 1 §5 d ^>^J 4,^J |!:^a IM s^s \ ^ ^^J^ ^lilc Q^l I" .^^< s^^ N 1 1 \ { ! '< 1 1 fifth day that the disease had spread to seven combs, the total number of diseased larva) being, as a whole, less than in the hybrid colony. There was not brood in all seven combs at the time of infection, but the brood increased faster than the disease spread. After the twenty-fourth day a permanent improvement began to be manifest. 12 BULLETIN" 804, U. S. DEPAETMENT OF AGRICULTURE This improvement continued after the queen was caged and became more marked after she was removed from the colony. These bees were better house cleaners as well; the appearance of larvse remaining over more than one observation period did not be- come evident until after the ninth day, compared with the sixth day in colony G. At no time were there as many of the larvae nearly ready to pupate that were gummy or rubbery. Even though this colony was on the average weaker than colony G all the time, it handled the disease much better. It was 14 days before colony G had cleaned up to such an extent that it was deemed safe to intro- duce a new queen, while in colony F, with the Italian bees, the combs were so nearly cleaned of everything but a few old scales that a five-frame nucleus with a new Italian laying queen was united with this colony after a 10-day queenless period and in 9 days more everything was absolutely clean and the queen was laying in the combs that had had disease in them. When an observation was made nine days after the new queen's eggs were first noted, it was found that there was a slight recurrence of disease in three of the combs. But, unfortunately, at the same time, queen cells and no eggs were found, denoting that for some reason this queen had not been accepted. Therefore the queen cells were all removed and a new queen was introduced. Although the author's observations ended of necessity soon thereafter, it was reported to him that this colony was doing nicely later in August and was perfectly healthy. If the first new queen had not disap- peared, it is quite probable that as soon as a sufficient number of her bees had emerged they would have cleaned up the recurring disease in the same manner as was done in colony J, which will be mentioned later. Several times in this colony, during the cleaning-up process, bees were watched in the act of sucking up juices of diseased larvae that had been partially removed from the cells with the aid of forceps. COLONY H Race. — Hybrid, a division of Colony G, hybrid. Queen. — 1918. Of their own raising. Poor. Bees. — Darli hybrids, almost black, excitable. Condition of colony at time of infection. — Brood in three frames, a few eggs in one, only one small patch sealed, the remainder from eggs up to 4-day larvse. Bees covering about live frames. Fairly good proportion of nurse bees. Date of first Infection. — July 1, 191S; second infection, July 8, 191S. Material used. — 20 old, dried, rubbery, diseased scales from sample No. 5898, macerated in 2.50 c. c, of a 50 per cent sugar sirup, colored with eosin. First appearance of disease noted. — July 5, doubtful. Positive July 8, 7 days after infection. Age of larrxr first attacked.— Fnur dnys after liatc-liiug from tlio ogg. BEES IN COLONIES AFFECTED BY EURC«>EAN FOULBROOD 18 Colony H (fig. 4) was treated as a double experiment. The infec- tion of this colony was not started until after the honey flow had come on quite heavily. Also, instead of freshly diseased larvae, old brown rubbery scales were used that showed Bncillus pluton present micro- scopically, but were heavily overgrown by Bacillus alvei. It was de- sired to learn whether these scales were still infectious, so that nurse bees working on them, cleaning them out, might get infective material on their feet and mouth parts which could be carried to healthy larvae. This was noted later in the observation hive, where, under the magni- fying glass, bees were seen trying to remove some of these rubbery- scales, first moistening them with their tongues and then pulling at them with the mandibles and front feet. This colony, which was marked hybrid and weak, was slow in developing the disease, partly because of the diluting effect of the z /-jv xs rtS ^£ >**^ '/V/ V£^ 0£S/!.(*f7X.V ,^f7i'Ck.> /^. 3** ' AV* ' CV£ O/t's^s^yA^ r/i o ra^'^^- ^ v^ /A oc s/f -■ ir/, ? 'J Q D . o CO o"tii EE D ^¥¥ • • • Q§: e?-- n £^ ••• / ^ ^ ^ ^ ^ P' <^ ^ /o // /^ /^ /^ /^ y^ yp'/id Fig. 6. — Distribution of cells containing infected sugar sirup and subsequent spread of the disease in a comb taken from colony H Area covered by brood at time of infection, mostly unsealed. • Loc-atlon of cells con- taining infected colored sugar sirup on July 3, 1918. ® First iwsitive diseased larvae noted, 2 on July 8. © Number of new diseased larvaj (4) on July 10. A Number of diseased larvae (13) on July 12. D Number of diseased larvije (39) on July IG. O Number of diseased larvae (52) on July 19. DISTRIBUTION OF INTRODUCED INFECTED MATERIAL An interesting experiment was carried out Avith sugar sirup, colored by a small amount of a harmless anilin dye, eosin, used as an indi- cator, which gave to the sirup a bright red color. The object of this experiment was to determine where the sirup, or, more important, where fresh nectar is first placed in the hive and combs. On May 27 two colonies were fed this colored sirup from above some time before the heavy honey flow from clover started. The results were striking, for in nearly every case the colored sirup was easily discernible in the cells and the greatest part of the sirup was located in quite a definite area. These colored cells were either scattered among the cells con- taining the larva) or were placed in a ring of cells adjacent to the brood area toward the top of the comb, little being placed with the solid stores (fig. 6). Furthermore, for nearly 36 hours after the feeding practically all the young nurse bees showed a marked pinkish discoloration of the anterior end of their abdomens, denoting the dis- BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD 21 tention of the honey stomachs by the retention of the colored sirup therein. About half the bees in the hives were discolored in this manner. After a day or so, however, this begin to disappear. Also the number of cells showing the pink discoloration began to disappear. Evidently the sirup had been moved up, worked over, and mixed with other nectar or consumed. Later, some time after the heavy honey flow had started, shortly after July 1, two more colonies were fed colored sirup, this time infected with diseased larvae macerated therein. In these cases the discolored abdomens were noted about as before, but the colored cells were less numerous and the color less striking. The location of the colored cells was similar to that in the former ex- periment ; that is, mainly in the brood area or just contiguous to it and mostly above. The outside combs, containing considerable honey, showed scarcely any of the colored cells. This time these colored cells disappeared sooner, showing that the infected material must have been much diluted quite soon after being taken up from the feeding dishes. Figure 6 shows the method of plotting the location of diseased larvae in the combs and also the location of the cells containing the colored sugar sirup. As will be noted, a fairly large proportion of these cells are located within the area of brood at the time of feed- ing. It is interesting to note the tendency of diseased brood to form concentric circles, showing the two series of larvae occurring between the dates noted. The spreading was from two cells at first to quite a large number at the last observation shown. AGE AT WHICH LARV^ ARE INFECTED In previous observations it was constantly noted that the larvae affected by European foulbrood were regularly at least 4 days old, the age at which the coiled larvae completely fill the bottom of the cells. Occasionally a slightly younger and smaller larva would become diseased, but this was not the common occurrence. Further- more, in the cases where the colored sirup was fed the bees, within 24 to 36 hours quite a number of larvae averaging 4 days old could be seen discolored from having been fed this sirup, while it was notice- able that the younger larvae under 3 days old never showed the dis- coloration. These colored larvae were examined in a smear under a microscope, but the infecting organisms, being comparatively few in number, had not increased sufficiently at that time to be apparent. The question now arises as to the age at which the larvae first are fed nectar or infected material. There has been much controversy over the subject of composition and source of the larval food, but as yet no conclusive scientific evidence has been presented. Irrespective of the question whether the food at various stages originates from glands or is regurgitated, it is apparent from these observations that 22 BULLETIN 804, V. S. DEPARTMENT OF AGRICULTURE there must be a difference between the food which larvae younger than approximately 3 days old receive and that fed to older ones. Otherwise the younger larvae w^ould also show^ the pink coloration. Von Planta (9) by chemical analyses, of questionable exactness, how- ever, makes a division in the feeding of the larvae at the age of 4 days, at which time the high protein and low sugar content change to lower protein and higher sugar content. These analyses would tend to coincide with the above observations, only it is probable that the change begins earlier. Additional data upon this subject are recorded in Tables II and III, although the observations were primarily for another purpose. In order to obtain further information relating to a possible difference in resistance to disease between Italian and hybrid bees, a careful record was made of the time when eggs were first noted in empty combs after the infection of the colony and when larvae first showed disease thereafter. In the case of comb Special No. 2, the eggs were laid by an Italian queen in a healthy colony and then placed in a diseased colony. Colonies F, A, and I were of Italian stock while colonies G, H, and J were hybrid. In the recurrence of disease all w^ere given new Italian queens. As has been mentioned before, as soon as the bees of the new Italian queens emerged in sufficient numbers the disease disappeared. Table II THE FIRST APPEARANCE OF DISEASE IN COMBS IN WHICH EGGS WERE LAID AFTER THE COLONY WAS INFECTED Colony and comb No. Number of days after eggs were first noted in comb. 1 2 3 4 5 6 7 8 9 10 n 12 13 14 G4 X G6 .::::::; X G7 X G8 X X F 3 '. F4 X P7 X H3 X X H5 A2 X X A 4 A5 X A 7 X A 8 X Spec. 2 X 13... X RECURRENCE OF DISEASE AFTER EGGS OF A NEW QUEEN WERE FIRST NOTED IN THE COMBS J 2 X X J 3 J4 X J5 X J 6a X J6b X J 7 X J8 X F3 X X X F4 F7 1 i BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD 23 Table III. — Average time, under various conditions, in which disease becomes apparent in a colony after infection imth European foulhrood. {Averages taken from Table II) Before the heavy honey flow. Colony G, hybrid Colony F, Italian Average of these two Days. During the heavy honey flow. Colony H, hybrid Colony A, Itahan Colony I, Italian Average of these three Days 9 9 8i5 Colony J, hybrid originally Colony F, Italian originally Average of these two Recurrence of disease, after treat- ment, dur- ing honey flow. Days. The data shown, particularly in Table III, tend to disprove the theory that Italian bees have a natural immunity ior resistance. If a larger number of observations could have been made, the variation would have appeared less. The effect of the honey flow is evident, however. When it is a question of the age at which the larvse are fed material that contains infection, these figures are significant. In the life history of the bee, 3 days are spent in the egg and from 5 to 6 days as larva before capping, making a period of 9 days in all. After 3 days in the egg and after having been fed predigested food for 3 days, with the additional 24 to 48 hour period of incubation, as was observed earlier in this paper, the larva ought to show disease from the fourth to the fifth day after hatching, or the seventh to eighth day of its existence, if Von Planta's assumption is correct. From actual observation this was found to be true and from observa- tion of the averages in Table III it is seen that the first appearance of disease occurs between the seventh and ninth days, varying with the conditions of the honey flow. Referring to Dr. Miller's theories, it is hard to believe that there is not plenty of highly infectious material left in the colony after a 5 or 6 day period of queenlessness. Aside from actual /observations of moist, yellow, melting larvae present more than 6 days after the 24 BULLETIN 804, U. s. DEPAETMENT OF AGRICULTURE queen has been removed, the juices of which the workers sucked up with avidity, the final eggs laid will be just at the stage where the dis- ease first appears ; that is, 3 to 4 days after hatching, at the end of a 6-day period. Furthermore, even though the nurse bees do not feed to healthy larvse the material that is taken up in cleaning out the cells in varying stages of decomposition, infection, even from scales, may be carried on the feet, mouth parts, and tongue, particularly, as was definitely shown with colony H, since these scales are in- fectious. The period of queenlessness and the consequent house cleaning are absolutely dependent on the strength of the colony. A strong colony cleans up rapidly, particularly after the introduction of the new queen in a cage plugged with candy. A weak colony, on the other hand, has not sufficient bees to clean even after complete introduction of a queen, and the disease sOiOn appears again. Under average conditions, therefore, it would appear unsafe to allow less than a 10-day period of queenlessness in treatment of European foul- brood. MICROSCOPICAL BACTERIOLOGICAL OBSERVATIONS A large number of microscopic examinations were made of larvae under various conditions for the positive presence of the characteris- tic groups of Bacillus yluton. These examinations were made mainly as a check on the gross observations of the first appearance of the disease. Cover glass smears were made of crushed larvae, stained with carbol fuchsin and mounted in Canada balsam. These examina- tions were made at regular intervals after the colonies were infected, larvae of all ages being examined. It was found in the smears of those larvae showing the first slightly abnormal symptoms that Bacillus pluton was the only organism present. This substantiates White's (12) observations that before the disease could be detected by gross examination, by a histological study of sections of larvae during the period of incubation it was demonstrated that " in the production of the disease Bacillus pluton was the first invader of the healthy larvae." As the disease advanced in the various colonies, observations were made of larvae in various stages of decomposition. The bacterial con- tent was found to vary with the change of appearance of the larvae during decomposition. The presence of these secondary invaders easily explains the atypical appearance of certain types of European foulbrood that heretofore have been very confusing to the bee- keeper. For a short time after the death of the larva, the color remains a moist, creamy-grayish yellow. This is during the period when Bacil- lus pluton and such occasional secondary invaders as Streptococcus apis or Bacterium eurydice and other organisms, which do not form BEES IN COLONIES AFFECTED BY EUROPEAN FOULBROOD 25 spores, are predominant as described by White (12) and McCray (4). Soon the putrefactive spore-forming organisms increase in number, Bacillus alvei^ being the one most commonly found. This is seen particularly in the case of the more mature larva?, which when dying extend more or less irregularly in the cells, becoming the gray- ish brown slimy masses which develop into the dark brown granular rubbery scales. This fact has been observed for a long time in the many samples which have been received for diagnosis. A partial description of these scales and of the presence of Bacillus alvei in them is given by McCray and White (5), but the experimental observations described in this paper added to diagnostic observations show that this condition is generally much more pronounced and common than described by these writers from laboratory observa- tions. The rapid increase and peculiar process of decomposition of Bacillus alvei, after the death of the larva, often to the exclusion of all other organisms, accounts for this abnormal appearance. In the case of American foulbrood, almost never is any other organism found associated with the disease but Bacillus larvae, the cause of the disease. This accounts for the constancy of the symptoms as compared with the variation of symptoms in European foulbrood w^here there may be several secondary invaders. Furthermore, in making the smears of the diseased larvae upon cover glasses, the peculiar whitish saclike extrusion of the larval in- testines was often noticed on crushing the larvae preparatory to smear- ing, which White (10) describes as a gross diagnostic character. When this sac was removed and smeared separately, it was always found to be heavily loaded with Bacillus pluton. Therefore it is safe to assume that the intestinal tract is the primary focus of infection, while the secondary putrefaction takes place mostly in the body tis- sues of the dead larva. Coincident with the microscopic examination of larvae, several ex- aminations were made of the contents of the ventriculus, rectum, and in a few cases of the honey stomach and mouth parts of bees. These bees were presumably nurse bees taken from diseased combs, some in the very act of sucking up the juices of dead diseased larvae. Al- though insufficient observations were made to give conclusive evidence, some interesting information was obtained. As may be seen from Table IV, the number of cases where Bacillus pluton or other organisms associated with infectious material were found in the intestinal contents is not very large. However, of more ^Bacillus alvei originally was supposed to be the primary cause of European foulbrood, but has been proved by White and others to be only a common secondary invader. Bacil- lus alvei has purely putrefactive functions. From its cultural and biochemical character- istics, Bacillus alvei apparently belongs to the common Bacillus subtilis (hay bacillviB) group of spore-forming organisms, all having mainly putrefactive functions. 26 BULLETIN 804, U. S. DEPARTMENT OF AGRICULTURE importance probably. Bacillus pluton was found in a smear made from the mouth parts of a nurse bee and also in the contents of the honey stomach of another. If these observations had been carried out systematically, instead of only casually, it is expected that much more positive data might have been obtained along these lines, owing to what is known already of the habits of house-cleaning bees working on diseased material. Table IV. — The results of the microscopic bacterial examination of tfie contents of the intestinal tracts of nurse bees taken from diseased colonies Microscopic findings. G. F. H. A. I. J. K. Total. 1 11 2 24 3 2 18 1 4 9 9 17 12 6 7 97 Bacillus alvei or doubtful Bacillus 11 1 SUMMARY AND CONCLUSIONS In arriving at the following conclusions an effort has been made to state them in a manner which will indicate the substantiation of previous observations made both in the laboratory and in the apiary. It may be noted that many of these conclusions are similar to some of the statements made in Farmers' Bulletin 975 in the summary of facts which apiary practice has brought out. 1. European foulbrood is an infectious disease. BaciUus pluton was found to be the primary invader, appearing in the intestinal tract of larvre before death, contemporary with the first slightly apparent symptoms. 2. The variation in the appearance of the diseased larvse after death is due to the presence or absence of secondary invaders. 3. The period of incubation for European foulbrood was found to be from 36 to 48 hours, although the gross symptoms usually do not become apparent in less than 3 or 4 days, varying with conditions of honey flow and strength of colony. 4. It has been noted in apiary practice that the first brood of the year usually escapes with little loss. During the first 5 to 7 days the spread of the disease in the colony after infection is slow, after which the increase is rapid under favorable conditions. The critical time, therefore, to detect the disease and start treatment is early in its course, thus making conditions unfavorable. 5. The evidence tends to confirm the theory that one of the ways the disease is spread in the colony is by the house-cleaning bees, and from colony to colony by their drifting. It is quite probable that the infective organisms are carried on the month parts and pedal appen- dages. The question of infection from intestinal contents or from BEES IN COLONIES AFFECTED BY EUEOPEAN FOULBROOD 27 the source of larval food at various stages needs further substantia- tion. 6. Irrespective of strength of colony, the Italian bees were found to resist infection much better than hybrids and showed more ability to overcome the disease. 7. This apparent resistance of the Italian bees was observed to be largely due to the more vigorous house-cleaning characteristics rather than to a natural resistance or immunity to the disease. There was very little difference in the apparent period of incubation between the Italian and hybrid colonies, possibly a slight difference in favor of the Italians. Furthermore, it was noted that often there may be a slight recurrence of disease in the brood of the new Italian queen until a sufficient number of her bees have emerged to eliminate the infection by house cleaning. Apparently, infection is not always entirely removed by a period of queenlessness. 8. As a rule, requeening is necessary in the treatment of European foulbrood, except possibly in the strongest Italian colonies, which show only slight infection. Wliere a considerable quantity of dis- ease is present, sufficient to require treatment, it was found unsafe to use a period of less than 10 days' queenlessness, due to the infec- tious condition of the diseased material remaining and the accom- panying behavior of the colony. 9. The stronger the colony in Italian bees, the more rapid was the recovery. . 10. A heavy honey flow tends to prevent infection from gaining a foothold. It also tends to eliminate the disease if present before the start of the heavy honey flow. This was found to be due to the effect of dilution on the infection because of the influx and direct feeding of the fresh nectar to the larvoe. 11. European foulbrood is a disease of weak colonies. It was found to be difficult effectually to infect any but the very weak colonies during the heavy honey flow. Therefore, colonies kept strong up to the time of the honey flow run very little danger of contracting European foulbrood. This and others of the facts ob- served are in exact harmony with facts already observed in apiary practice. LITERATURE CITED (1) Alexander, E. W. 1905. How to rid your apiary of blacli brood/ In Gleanings in Bee- Culture, V. 33, p. 1125. (2) Miller, C. C. 1911. Fifty years among the bees. (3) 1918. European foulbrood and its treatment. In American Bee Journ., V. 58, no. 7, p. 232-234. July. (4) McCray, a. H. 1917. Spore-forming bacteria of the apiary. In Jour. Agr. Research, V. 8, no. 11, p. 399-420, pi. 93-94. (5) and White, G. F. 1918. The diagnosis of bee diseases by laboratory methods. U. S. Dept. Agr. Bui. 671, 15 p., 2 pi. (6) Phillips, E. F. 1911. The treatment of bee diseases. U. S. Dept. Agr. Farmers' Bui. 442. May 6. (7) 1918. The control of European foulbrood. U. S. Dept. Agr. Farmers' Bui. 975. July. (8) [Root, E. R.] 1905. (Editorial.) In Gleanings in Bee-Culture, v. 33, p. 1126. Nov. 1. (9) VoN Planta, a. 1888. Ueber den Futtersaft der Bienen. In Zeit. f. Pliys. Chemie von Hoppe-Seyler, v. 12, p. 327-354. (10) 1889. Ueber den Futtersaft der Bienen. In Zeit. f. Phys. Chemie von Hoppe-Seyler, v. 13, p. 552-561. (11) West, N. D. 1899. Foul and other forms of diseased brood in the State of New York, In Gleanings in Bee-Culture, v. 27, p. 828. Nov. 15. (12) White, G. F. 1912. The cause of European foulbrood. Cir. 157, Bur. Ent., U. S. Dept. Agr. May 10. ' " Black brood " is an old name for European foulbrood. 28 ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D, C. AT 5 CENTS PER COPY 5«,p,f J'. UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 805 Contrtbatlon from the BoreBU of Entomology L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER December 15, 1919 TWO LEAFHOPPERS mJURIOUS TO APPLE NURSERY STOCK Br A. J. ACEERMAN, Sdentiflc Assistant Deciduous Fruit Insect Inyestigations CONTENTS Page Introduction 1 The Apple Leaf hopper . * 2 History 2 Distribution 2 Food Plants 3 Cltaracter of Injury 4 Extent of Injury and Influencing Factors 5 Description of Stages 5 Allied Species 6 Life History and Habits 7 Summary of Seasonal Ifflstory . . 19 Natwal Enemies 20 Page The Rose Leaf hopper ° . . 20 History 20 Synonymy 21 Origin and Distribution 21 Food Plants 21 Character of Injury 22 Description of Stages 22 Life History and Habits 23 Summary of Seasonal History . . 28 Natural Enemies 28 Remedial Measures 29 Summary 83 Literature Cited M WASHINGTON GOVERNMENT PRINTINO OFFICE 1919 HB UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 805 Contribution from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER December 15, 1919 TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. By A. J. ACKERMAN, Scientific Assistant, Deciduous Fruit Insect Investigations. CONTENTS. Page. Introduction 1 The apple leafhopper 2 History 2 Distribution 2 Food plants 3 Character of injury 4 Extent of injury and influencing factors 5 Description of stages 5 Allied specie<5 6 Life history and habits 7 Summary of seasonal history 19 Natural enemies 20 Page. The rose leafhopper 20 History 20 Synonymy 21 Origin and distribution 21 Food plants 21 Character of injury 22 Description of stages 22 Life history and habits 23 Summary of seasonal history 28 Natural enemies 28 Remedial measures 29 Summary 33 ^ Literature cited 34 INTRODUCTION. Serious injury to apple nursery stock due to the attack of leaf- hoppers attracted the writer's attention while engaged in the investi- gation of nursery fruit insects at West Chester, Pa. An examina- tion of the injury showed the presence of two species of leaf hoppers, the common apple leafhopper^ and the rose leafhopper.^ Further study proved that each species produced a distinct type of injury, that caused by the apple leafhopper being by far the more serious. The conflicting nature of the entomological literature regarding the char- acter of injury caused by these two species and their habits led the writer, under the direction of Dr. A. L. Quaintance, of the Bureau of Entomology, to undertake a study of their individual life histories and the means for their control. 1 Empoasca mali (Le Baron); order Hemiptera, suborder Homoptera, family Cicadellidae. ' Empoa rosne (Linn.); order Hemiptera, suborder Homoptera, family Cicadellidae. 132816°— 19 1 2 BULLETIN 805, IT. S. DEPARTMENT OF AGRICULTUKE. The destructiveness, habits, food plants, and life history of each species are treated separately herein. A detailed account of the ex- perimental work carried out for the control of these two species to- gether with the niost efficient remedy is included. The data on biology and control were obtained at West Chester, Pa., during the seasons of 1915 and 1916, supplemented by field ob- servations at several points in southeastern Pennsylvania, and western Maryland. THE APPLE LEAFHOPPER, Empoasca mali (Le Baron). HISTORY. This species was originally described by Le Baron (1) in 1853 ^ under the name Tettigonia mali, and it was recorded by him as in- jurious to fruit trees in Illinois. In 1862 the genus Empoasca was erected by Walsh (2) with a description of three new species, but no mention was made of mali. Carlos Berg (3, p. 273), in 1879, described a leaf hopper from Argentina as TypTilocyha phytophila, and this name later was considered by Gillette to be a synonym of Empoasca mali. In 1883 Forbes (4) sent specimens of a green apple leaf hopper to Uhler who determined them as belonging to the genus Empoa; subsequently these insects were described by Forbes (5) as a new species, Empoa alhopida. Woodworth (6) transferred Empoa albopicta Forbes to the genus Empoasca in 1889 and called it Empoasca albopicta. The first reference to this species under its correct name, Empoasca mali, was made by Gillette (7) in 1890. Osborn (8) and later Gibson (17) mentioned it as injurious to potatoes, and Gillette (9), in 1898, gave the food plants and distribution. Frequent references to this insect have been made in American entomological literature under the name of "the apple leafhopper" and "the currant leafhopper," by Britton (12), Brues (13), and Gar- man (16) among others. It has been often referred to as the most injurious leafhopper, both to apple and to various field crops. Wash- burn (15), in 1908, was the first writer to treat of this insect at any length. He published a record of the seasonal history, food plants, injury, and control of the apple leafhopper as a nursery pest in Min- nesota. In 1910 K. L. Webster (18) made a detailed study of the life history and control of this insect on apple nursery stock in Iowa. In 1915 Webster (20) published an account of Empoasca mali, treating it as a pest of potatoes. DISTRIBUTION. There are no records showing that the apple leafhopper occurs in Europe. In America, outside of the United States, it has been re- ported from Okanagan, British Columbia, from Nova Scotia, several 1 Figures in parenthesas refer to "Literature cited," p. 34. TWO LEATHOPPERS INJURIOUS TO APPLE NURSERY STOCK. points in the Province oi Ontario, from Mexico, Porto Rico, and Cor- rientes, Argentina. In the United States this species is widely dis- seminated, doubtless due to the variety and abundance of its host plants. From speci- mens in the collection of the United States National Museum, and from the collec- tion, correspondence, and notes of the Bu- reau of Entomology, it appears to be pres- ent in almost every State in the Union. (See fig. 1.) It is found in greatest abundance throughout the east- ern humid area of the Upper Austral Zone. FOOD PLANTS. The food plants of Empoasca mali (Le B.) are very numerous and varied. In nurseries this insect prefers apple but it also feeds in great abundance on Norway maple and various oaks. Among field crops it is partial to alfalfa, clover, potato, and beets, in about the order named. A list of all host plants reported, upon the majority of which the writer has noted this insect feeding, follows : Fig Distribution of the apple leafhopper (Empoasca mali) in the United States. Acer negundo, box-elder. Acer platanoides, Norway maple. Althea rosea, liollyhock Amygdalus persica, peach. Apium graveolens, celery. Avena sativa, oats. Beta vulgaris, beets. Betulaep., birch. Cannabis sativa, hemp. Castaneasj)., chestnut. Corylus amcricana, hazelnut. Crataegus sp., hawthorn. Cydonia oblonga, quince. Dahlia STp., dahlia. Gramineae, grasses. Hamamelis virginiana, witch-hazel. Hicoria pecan, pecan. Juglans nigra, black walnut. Juglans sp., walnut. Medicago sativa, alfalfa. Phaseolus vulgaris, beans. Populus sp., poplar. Prunus virgianiana, choke-cherry. Prunus pissardi, purple-leaved plum. Prunus spp., cherries and plums. Pyrus baccata, Siberian crab. Pyrus communis, pear. Pyrus malus, apple. Quercus spp., oaks. Rfieum rhaponticum, rhubarb. Rhu^ cotinus, smoke-tree. Ribes oxyacanthoides, gooseberry. Ribes rubrum, currant. Rubus spp., blackberry and raspberry. Rosa spp., roses. Secale cereale, rye. Solanum tuberosum, potato. Sorbus amcricana, mountain ash. Sorghum sp., sorghuln. Syringa 8T[>., lilac. Tilia amcricana, American linden. Trifolium sp., clover. Uhnu^ amcricana, American elm. Viburnwn sp., snowball. Vitis spp., grapes. Zea sp., com. 4 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. CHARACTER OP INJURY. The injury caused by the apple leafhopper to nursery apple trees is due to the feeding of the nymphs and adults on the underside of the tender terminal leaves from which they extract the plant juices. As a result of this attack the leaves become undersized and curled (PI. I, B), causing a decided check to the growth of the new wood. The curling begins at the apex and extends toward the base of the eaves, the lower surface always being rolled in. Tliis type of injury differs from aphis leaf-curl in that aphids roll the leaves more tightly and curl them from the sides instead of from the tips. During the progress of the injury produced by Empoasca mdli the leaves become wrinkled and the loss of sap finally causes the tips to dry up and turn brown. (See PI. II, fig. 2.) The nymphs, because of their greater numbers and due to the fact that they spend the entire nymphal period on a few leaves only, cause more serious injury than do the adults. The latter feed only for a short time, being principally engaged in egg-laying, and during this period they fly from one tree to another. Injury by the feeding of the adults, therefore, is of little importance when compared with the local- ized injury produced by the nymphs. Consequently, the stunted ter- minal growth is most apparent at the time when the nymphs are most abundant on the foliage. As the nymphs gradually disappear the terminal shoots seem to revive and develop normal leaves above the stunted ones. (See PI. I, A.) At the time of infestation by the next brood of nymphs, however, a similar check to the new terminal growth is produced. Thus retardation in growth occurs periodically throughout the season corresponding to the periods of infestation by the successive nymphal broods, while intervening between each infes- tation there is a short period during which the terminals maintain a normal growth. Although the different broods of nymphs over- lap slightly the successive checks in terminal growth usually are well defined. In the vicinity of West Chester, Pa., there are three broods during the season and three corresponding checks in the terminal growth. The first growth-check takes place during the latter part of June when the first nymphal brood is feeding; a second and a third check appear during the latter part of July and August, respectively, at the tune when the second and third broods of nymphs are most active on the foliage. The first brood is the most abundant on apple and consequently causes more injury than do either of the two fol- lowing broods. Adults of the first brood do not confine their activ- ities to apple alone, as many scatter to other host plants to feed and oviposit. Bui. 805, U. S. Dept. of Agriculture. Plate I, The Apple Leafhopper (Empoasca mald. A, Terminal leaves of apple shoot outgrowing injury by the leafhopper; B, curled condition of terminal leaves caused by the leafhopper. Bui. 805, U. S. Dept. of Agriculture. Plate 1 1 , Fig. I.-Cages Used for Rearing the Apple Leafhopper in Nursery, West Chester, Pa. ■■^-•^^ '^-l Fig. 2.— Terminal Leaves of Nursery Apple Trees Curled by the Apple Leafhopper. THE APPLE LEAFHOPPER. Bui. 805, U. S. Dept. of Agriculture. PLATE III. The Apple Leafhopper. A, First nvinphal stage; B, second .stage; (', third stage; D, fourtli stage; E, fiftli stage; F, side view of fiflli stage; G, adult; //, front view of liead of adult; /, eggs in tissue on underside of apple leaf; J, curled condition of terminal leaves due to attack by the apple leafhopper on apple. TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 5 As a result of the contimied checking of the growth, due to the in- festation of the apple leafhopper, nursery apple trees often requii'e an additional year's growth before they become of marketable size. EXTENT OF INJURY AND INFLUENCING FACTORS. The extent of injury varies according to the age of the nursery- stock and according to the differences in the character of growth of apple varieties. Seedlings and the initial growth of buds and grafts are very seriously injured. Nursery stock at this stage is in its most critical period of growth and is injured very easily. Furthermore, any injury at this stage is not readily outgrown. After the first year's growth the more vigorous varieties become partially immune to serious injury and succeed in maintaining a satisfactory growth, while slow-growing and tender-leaved varieties are at all times badly injured by the attack of tliis insect. This is easily understood since, even under normal conditions, the latter make but a very ordinary growth and are entirely unable to with- stand a serious check. Among the varieties most severely injured in Pennsylvania nurseries Red Astrachan, a particularly slow grower during the first two seasons, ranks first, followed by Smith's Cider, Starr, Early Harvest, Summer Rambo, Delaware Winter, Wagoner, Golden Russet, Early Ripe, Wealthy, and Alexander. description of stages. Egg. ri. Ill, I. The egg is elongate, subcylindrical in form, very delicate, slightly curved from end to end, somewhat rounded at both ends but more so at the anterior one. When first deposited it is rather transparent but in a few days it changes to a pale yellow color, wliile a small white cap forms at the anterior end through which the red eyes of the immatiure nymph are perceptible. Average length of 15 eggs 0.82 mm., width 0.25 mm. Nymph. PI. Ill, A-F. First instar. — Color pale white, changing to a light yellowish green after feeding. Eyes dull red. Small pale spines on the dorsal side of the head, thorax, and abdomen; the latter with four spines to each segment arranged in two longitudinal rows along each side, one spine situated dorso,laterally, the other ventro-laterally. Posterior margin of metathorax blunt. First two segments of antennae pale, the remainder dusky. Average length of 16 specimens 1 mm. Second instar. — General color Hght yellowish green. Eyes lose some of their red color. Posterior border of metathorax sharp in outline. First two segments of antennae light yellow, remainder dusky. Average length of 16 specimens 1.30 mm. Third instar. — General color pale yellowish green. Eyes almost pearl white. Body more robust than in first two stages. Wing pads appear as lateral buds extending 6 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. to the Mud mai^n of the first abdominal segment. Spines darker and more prominent. Average length of 16 specimens 1.85 mm. Fourth instar. — Head and thorax yellowish green; abdomen yellow in color. Eyes pearl white. Wing pads extend to hind margin of second abdominal segment. Spines prominent. Average length of 16 specimens 2.1 mm. Fifth instar. — Head and thorax pale green; abdomen yellow. Eyes dull white. Wing pads extend to or nearly to the hind margin of the fourth abdominal seg- ment. First two antennal segments green, remainder dusky. Body broader than in previous stage. Average length of 16 specimens 2.6 mm. Adult. PI. Ill, G, H. General color of adult pale green; face with a white median longitudinal Une in older specimens but composed of a series of white spots in newly hatched individuals; median line extending from a point midway between the ocelli to a point half the distance to lower margin of clypeus; two short white diagonal bands on each side of median line, the lower one the smaller; a short white line, often merely two spots, beyond the diagonal and just above the antenna; a faint white line midway between the ocellus and eye; antennae 1 mm. in length, arising near the lower frontal border of the eyes; clypeus one-third longer than broad ; lorse narrow, not reaching the tips of clypeus, con- cave below eyes; genaj almost as long and half as broad as clypeus, with one or two faint white spots. Vertex dark green with a median white line, narrowest in middle, its length equal to distance between the ocelli ; a white band on each side, dorso-lateral and diagonal to median line. Two ocelli present, marked by two white spots and situated on frontal margin of vertex, their distance apart equal to twice that from the eye to the ocellus; eyes dull wliite, reddish brown after death. Pronotum pale green, hind margin very pale, with eight white spots along the frontal margin, the last spot at each end small and often fused with the one next to it so as to form only six spots; mesonotum with two parallel white longitudinal lines centrally located and con- nected by a traverse one in the form of a letter H , a faint white diagonal Hne present on each lateral margin; scutellum small with a large white triangular area in the center and a small spot on each side along the frontal maig;in. Abdominal segments yellow- ish green with transverse yellow stripes on their liind margins, anal segment dark green. Wings semitransparent, pale yellowish green. Legs green, tarsi dusky at the tips. Sexual appendages ciliated in both sexes. Average length of 16 specimens 3.12 mm. ALLIED SPECIES. Three other species of Empoasca were found associated with E. 7nali on the foliage of nursery apple trees at West Chester, Pa. These species were determined by the late Otto Heidemann, of the Bureau of Entomology, as follows: E. birdii Godmg, E. jlavescens (Fabricius), and E. unicolor Gdlette. Birdii Siud jlavescens are very closely allied species, the former being considered by Gillette to be merely a color variety of the latter; these two species resemble mali quite closely and they may be easUy mistaken for it. Unicolor, on the other hand, differs markedly from any of the above three and is readily distinguished from them. Birdii differs from mali by its smaller size and paler color, by the presence of smoky markings on the elytra, and by the three white spots on the pronotum. No attempt was made to study the life history and habits of hirdii but they are probably much the same as those of mali. From TWO LEAFHOPPERS INJtfRIOUS TO APPLE NURSERY STOCK. 7 field observations it was found that this species hibernates in the adult stage in woodlands near the nursery at West Chester. The adults become active in the spring about a week earlier than mali, confining their feeding at first to the foliage of skunk cabbage. From this plant they scatter to grasses and weeds beneath the apple trees in the nursery a few days prior to the first appearance of mali. During the early spring they prefer to feed on any low green vegeta- tion in the nursery row, and never become abundant on the foliage of apple until about mid-season. At this time they appear m num- bers associated with mali on the terminal leaves. The extent of damage caused by hirdii is small compared with that caused by the common apple leaf hopper, and for this reason little attention has been paid to it heretofore. Flavescens differs from hirdii by the absence of the characteristic white markings of the pronotum and the smoky bands crossing the elytra. This species is allied very closely to hirdii m appearance, and proba- bly in habits and life history, but it is less abundant on the apple. Unicolor is readily separated from mali by the absence of the conspicuous white markings of the face and the notum, by its greater length and robustness, and by the presence of a pale white spot on the middle of the anterior margin of the pronotum and a blue blotch on the scutellum. Few field observersations were obtained in regard to the habits of unicolor. The nymphs of the first brood were found on apple at approximately the same date and about as abundantly as those of ttmU at Hagerstown, Md. The adults of this species do not confine their attack to the termuial leaves, being found more frequently on the lower part of the trees. This species was taken in scant numbers on apple in the vicinity of West Chester and the injury caused by it was negligible. LIFE HISTORY AND HABITS. Methods op Study. In studying the life history of the two species of apple leafhoppers concerned in this bulletin all data were ob tamed under outdoor conditions by rearing the insects on young apple stock in the nursery row. Seedlings were planted out early in the spring of 1915 and again in 1916 on a plot of ground at one end of a few nursery rows. Riley cages and arc-light globe cages (PI. II, A) were used for obtaining records of the length of the egg stage, the extent of repro- duction per female, and the longevity of adults of the different broods. The globe cages were well shaded from the sun by means of large muslin covers over their tops, while ventilation was obtained both from above and below. With the use of such cages practi- cally normal conditions were secured for the rearing of the leaf hopper 8 BULLETIN 805, V. S. DEPARTMENT OF AGRICULTURE. material. The plants were encaged before the hibernating adults made their appearance in the nursery, thereby preventing outside infestation. Special cages were constructed for experiments in determining the length of the nymphal stages. Various types of cages were tried in an effort to secure one in which the nymph could be reared under as nearly normal conditions as possible. The type of cage finally decided upon was made as follows: A piece of thin sheet-cork was cut about 2 inches square, in the cen- ter of which a 1-inch square hole was made. White muslin cloth was stretched tightly over one side of the cork and glued fast so as to cover the center hole. Heavy waddmg cut to the shape of the cork, but leaving the center open, was glued to the other side. With the muslin side out, the cage was then placed over a newly hatched nymph on the lower surface of a leaf. A square of stiff cardboard of the same size as the cork was placed on the upper side of the leaf, and the cardboard, leaf, and cage were fastened together by paper clips. The young nymph within the cage received ventilation from both sides, through the porous wadding and through the muslin top. The leaf tissue was protected agamst injury by the cardboard on its upper surface and by the wadding on its lower surface. The nymph was examined daily by removing the paper clips and lifting the cage slightly; in this manner a record of the molts was obtained. Although this cage was a little heavy when used on the small leaves of seedlings, it proved satisfactory when fastened to the larger leaves of two-year trees. For this reason the nymphal stages were obtained by transferring newly hatched nymphs from globe cages to the cork cages on uninfested leaves of older trees in the nursery row. Number op Generations. There are three generations of the apple leafhopper at West Chester, Pa. These generations overlap slightly but they are easily distin- guished by the resultant injury caused by each. The first generation, covering the period from the time of egg deposition by the overwin- tered females to the death of the first-brood adults, extends from the last week in May to the first week in August. The second generation covers the period from the first week in July until the latter part of September. The third generation, including the hibernating adults, lasts from the first week in September until the early part of July of the following season. Adults of this generation hatch during the first week in September and remain on the trees until late in November when they seek shelter for the whiter. In the spring of the following year overwintered adults are found on the trees from the last week in May until death, which occurs during June and early July. TWO LEAFHOPPEES INJURIOUS TO APPLE NURSERY STOCK. 9 Forbes and Hart (10) in 1900 mentioned the occurrence of four or more generations in Illinois. In 1908 Washburn (15) suggested that there were two and possibly three generations in Minnesota. R. L. Webster (18) in 1910 recorded four generations at Ames, Iowa. E. H. Gibson, of the Bureau of Entomology, has noted as many as five generations in southern Missouri, and six m southern Mississippi. The Egg. The eggs are laid singly in the sides of the mid-vein and occasionally in the smaller veins of the terminal leaves. They are deposited in pockets just under the epidermis, usually lying in a longitudhial position. It is very difficult to locate the eggs as they are the same color as the tissue in which they are embedded, while the epidermis rnider which they are hidden is covered by the pubescence of the leaf. When the pubescence is removed, the tissue covering the egg appears slightly distorted and eventually becomes discolored. In making dissections of the leaf tissue the delicate egg is often crushed, where- upon the egg contents may be mistaken for the plant juice in the vein. When ready to hatch, the immature nymph pushes its head through the anterior end of the eggshell and forces a tiny hole in the thin epidermal leaf-covering, slowly drawing its body free from the enveloping tissue. Eggs of Empoasca mali have been found in the leaves of the follow- ing host plants: Apple, pear, peach, plum, cherry, quince, alfalfa, beet, and potato. Adults of all three generations deposit summer eggs in leaves in the manner mentioned above. Washburn (15) stated in 1908 that the last-brood adults of this species deposit winter eggs under the bark of nursery apple trees in Minnesota, and that the nymphs hatch- ing therefrom the foUowmg sprmg attack the lower leaves of the trees. Webster (18) in 1910 made similar observations in Iowa. At West Chester, Pa., the apple leafhopper certainly does not pass the winter in the egg stage. Several experiments were made in the attempt to obtain winter eggs by confining numerous pairs of third-brood adults in cages, but all proved unsuccessful. Field observations for two seasons on several thousand trees also substantiate the above view. However, winter eggs of the rose leafhopper (which will be treated later) were found in abundance in this locality, the nymphs of which confine their feeding to the lower leaves of the trees. The Nymph. The newly-hatched nymphs are very small, wingless, white in color, and of the same form as the adults. Immediately after hatching they settle down to feed, inserting their minute beaks in the leaf tissue and sucking the plant juices. A day or. two after taking food into 132816°— 19 2 10 BULLETIN 805, IT. S. DEPARTMENT OF AGRICULTURE. their bodies the young nymphs change to a pale green color, which ia the characteristic color during the remainder of their nymphal life. The nymphs pass through five stages of development before they reach maturity, molting and increasing in size at the completion of each stage. The nymphs are very agile in their movements and run in a zig-zag or sidling manner; only fourth and fifth stage nymphs are able to hop. Tlie first nymphs of the season appear on the trees about June 1, and the nymphal infestation is at its height about tliree weeks later. The Adult. The adults are very active, especially on warm, sunny days, when they rise from the trees in swarms at the least disturbance. During fhght the hoppers seldom rise over the tops of the trees but fly sidewise to the next nursery row. Records have appeared stating that this insect is strongly attracted to artificial light, but this view is contrary to observations made by the writer. Among the leafhoppers found swarming around electric lights the two allied species E. Jlavescens and E. hirdii were far more abundant than this species. The overwintered adult leafhoppers become active during the warm days about the middle of May. Upon leaving their hibernating places they immediately disperse to apple blocks in the nursery. The first adults of the season appearing on apple stock in the nursery were found on May 18; by June 1 they become quite abundant on the foliage in this section. For several days they confine their activities to feeding on the underside of the terminal leaves. After feeding for about a week the adults mate and soon begin depositing the first- brood eggs. Adults of all three generations have the same habits on apple. Third-brood adults feed on apple foliage in the fall until cold weather sets in. Prior to going into hibernation they collect on the lower leaves of the trees and on several varieties of low-growing weeds in the nursery row, being especially abundant on sorrel (Rumex sp.). Longevity op Overwintered Adults. Data on the length of life of the adult apple leafhoppers were ob- tained with great difficulty, due to their activity. A Riley cage was used in this experiment in which 50 adults were confined on a young apple tree. The newly transforming nymphs were removed from the foliage from time to time to prevent confusing them with the adults. During the progress of the examinations a few adults made their escape. A record of the date on which the adults died is given in Table I. It will be noted that death occurred in most cases by July 1. From observations in the field it was found that practically TWO LEAFHOPPEKS INJURIOUS TO APPLE NURSERY STOCK. 11 all overwintered adults had disap|peared from the trees by the first week in July. Table I. — Longevity of overwintered adult apple leafhoppers; 50 adults placed in cage June 8, 1915. Date of examination. June 20 June 21 June 22 Do. June 23, June 24. June 25. June 26. June 27. Do. June 28. June 29. Number dead. Date of examination. June 30. Do.. July 1.. Julys.. July 12. July 13. July 14. July 15. July 16. July 17. July 18. July 19. Number dead. 1 Escaped. Extent op Reproduction by Overwintered Females. Pairs of overwintered adults were placed on young trees in globe cages to determine the number of eggs deposited by a female in con- finement. The adult leafhoppers were allowed to remain on the foliage untU they died. Examinations were made from time to time and a record was kept of the number of nymphs removed at each examination. This method of obtaining the extent of reproduction does not take into consideration the number of eggs that failed to hatch; however, it shows approximately the number of eggs deposited by a single female. Two experiments were carried out and averages of 27 and 28 eggs per female hatched in the respective cages. In Table II the number of nymphs removed is given. Table II. — Number of first-brood nymphs produced by overwintered adult apple leaf- hoppers in confinement, West Chester, Pa., 1915. Cage No. 1.— TWO PAIRS OF OVERWINTERED ADULTS PLACED IN CAGE ON YOUNG APPLE TREE JUNE 9, 1915. Dat« of examination. June 24 June 26 Jiine 29 Julyl.. Number of flrst- brood nymphs removed. Date of examination. July 7.. July 30, Number of first- brood nymphs removed. Cage No. 2.— THREE PAIRS OF OVERWINTERED ADULTS PLACED IN CAGE ON YOUNG APPLE TREE JUNE 7, 1915. 12 BULLETIN 805^ U. S. DEPARTMENT OP AGRICULTURE. FiKST Generation. FIRST BROOD OF EGGS. Length of incubation period. — The adult females are very rarely seen depositing eggs since they are very agile in their movements and take flight at the slightest disturbance. Moreover, even though a female is observed in the act of oviposition, the egg is seldom detected, as it is inserted under the epidermis and is further hidden by the downy pubescence of the leaf. For these reasons only approxi- mate records of the length of the egg stage could be secured. Data on the incubation period were obtained by confining copu- lating pairs of overwintered adults on a seedling tree in a globe cage. The adults were removed after 24 hours and a daily record was kept of the nymphs as they hatched. The length of the egg stage, together with the average mean temperature for the entire period, is given in Table III. The average incubation period was 7.5 to 9.5 days; the maximum, 10 to 12 days; the minimum, 5 to 7 days. Table III. — Length of incubation period of first-brood eggs deposited by overwintered females of the apple leaf hopper in confinement, West Chester, Pa., 1915. TWENTY PAIRS OF OVERWINTERED ADULTS PLACED ON APPLE TREE JUNE 13, 1915, 12 M.; ADULTS REMOVED JUNE 14, 12 M. Date of removal of newly hatched nymphs. Number of nymphs removed. Length of in- cubation period. Average tempera- ture for entire period. June 20, 12 m . . June 21, 12 m.. June 22, 12 m . . June 23, 12 m.. June 24, 12m.. Jime25, 12 m.. Average. Days. 5-7 6-8 7-9 8 -10 9 -11 10 -12 "F. 79.56 78.61 77.00 75.90 75.20 75.07 7. 5- 9. ; 76.89 FIRST BROOD OP NYMPHS. Length of stages. — The length of the nymphal stages and of the entire nymphal period was obtained by confining newly hatched nymphs in the cork cages described on page 8. Each cage was examined daily and a record made of the length of each stage, the cast skins being removed from the cages after each molt. The length of each stage and of the entire nymphal period is shown in Table IV; the average mean temperature for each nymphal period is also included. The average length of the nymphal period for 22 nymphs was 18.7 days; the maximum, 22 days; the minimum, 15 days. TWO LEAFHOPPERS INJURIOUS TO APPLR NURSERY STOCK. 13 Table IV. — Length of nymphal stages of the first brood of the apple leaf hopper at West Chester, Pa., 1915. First brood, June and July. Date of hatch- ing. First molt. First stage. Second molt. Sec- ond stage. Third molt. Third stage. Fourth molt. Fourth stage. Fifth molt. Fifth stage Total nym- phal pe- riod. Aver- age mean tem- pera- ture for nym- phal pe- riod. 1915. June 12 Do... Do... Do... June 13 Do... Do... Do... Do... June 14 Do... Do... Do... Do... June 15 Do... Do... Do... June 18 Do... June 19 June 23 1915. June 15 June 16 June 18 June 16 ...do.... June 17 June 18 June 16 June 18 June 16 June 17 June 18 ...do.... June 17 June 18 ...do.... June 19 ...do.... June 21 June 20 June 21 June 26 Days. 3 4 6 4 3 4 5 3 5 2 3 4 4 3 3 3 4 4 3 2 2 3 1915. June 17 June 18 June 21 June 19 June 20 ...do ...do ...do ...do June 19 June 20 ...do Jime 21 ...do ...do ...do ...do June 22 June 26 June 24 ...do June 30 Days 2 2 3 3 4 3 2 4 2 3 3 2 3 4 3 3 2 3 5 4 3 4 1915. June 21 ...do.... June 25 June 22 June 24 June 21 ...do June 25 June 23 June 21 June 24 June 23 June 20 June 23 June 24 June 25 June 24 June 27 June 29 ...do June 26 July 4 Days. 4 3 4 3 4 1 1 5 3 2 4 3 8 2 3 4 3 5 3 5 2 4 1915. June 25 June 26 June 29 June 27 ...do June 26 June 27 June 29 June 27 June 29 June 28 ...do July 2 June 26 June 28 June 29 June 28 July 1 July 3 July 2 June 29 July 8 Days. 4 5 4 5 3 5 6 4 4 8 4 5 3 3 4 4 4 4 4 3 3 4 4. 2 1915. June 30 July 1 July 3 July 1 July 2 July 1 July 4 July 3 July 2 July 3 July 2 July 3 July 6 July 1 July 3 July 4 July 1 July 6 ...do.... ...do... July 4 July 11 Days. 5 5 4 4 5 5 7 4 5 4 4 5 4 5 5 5 3 5 3 4 5 3 Days. 18 19 21 19 19 18 21 20 19 19 18 19 22 17 18 19 16 21 18 18 15 18 "F. 75.18 75.47 75.79 75.47 75.20 75. 15 75.70 75.52 75.20 75.20 74.84 75.20 75.28 74.77 74.81 75.05 74.32 74.95 74.36 74.36 74.06 75.07 Average 3.5 3 3.4 4.5 18 7 75 04 FIRST BROOD OF ADULTS. Longevity. — ^First-brood adults were corLfined in a Kiley cage to determine the length of life. Most of the hoppers died by August 1, as shown in Table V. Table V. — Longevity of first-brood adult apple leaf hoppers; 50 adults placed in cage July 2, 1915. Date of examination. July 7. July 8. July 9. July 10 Do July 11, July 12, July 13, July 17, July 24, Number dead. Date of examination. July 24 July 26, July 27 July 30, July 31, Aug. 2. Aug. 4. Aug. 5. Do Aug. 6 Number dead. 1 2 2 3 4 1 3 2 1 1 1 1 Escaped. Extent of reproduction. — Two globe cage experiments were made to determine the extent of reproduction by a first-brood female in con- finement. In one cage two pairs of adults, in the other three pairs were confined until death occurred. The average number of eggs 14 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. deposited per female in cage 1 was approximately 24; in cage 2 approximately 22. (See Table VI.) Table VI. — Number of second-brood nymphs jrroduced by first-brood adult apple leaf- hoppers in confinement, West Chester, Pa., 1915. Cage No. 1.— TWO PAIRS OF FIRST-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE JUNE 30, 1915. July 13. July 17. July 20. July 22. July 24. Date of examination. Number of second- brood nymphs removed. Date of examination. July 27.. July 30.. Aug. 10. Average. Number of second- brood nymphs removed. Cage No. 2.— THREE PAIRS OF FIRST-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE JUNE 30, 1915. Second Generation. SECOND BROOD OF EGGS. Length of incubation 'period. — ^As will be noted in Table VII, the length of the egg stage of this brood is somewhat shorter than that of the first brood. This fact undoubtedly is due to the higher tem- peratures occurring throughout the incubation period of the second- brood eggs. As in Table III only an approximation of the length of the incubation period was obtained. The average incubation period of second-brood eggs was 6 to 8 days; the maximum, 9 to 11 days; the minimum, 3 to 5 days. Table VII. — Length of incubation period of second-brood eggs of the apple leafhopper, West Chester, Pa., 1915. TWENTY-FIVE PAIRS OF FIRST-BROOD ADULTS PLACED ON APPLE TREE JULY 7. 1915, 3 P. M.; ADULTS REMOVED JULY 8, 3 P. M. Date of removal of newly hatched nymphs. Number of nymphs removed. Length of incu- bation period. Average tempera- ture for entire period. July 12,3 p.m. July 13, 3p.m. July 14, 3p.m. July 15, 3 p. m. July 16, 3 p. m. July 17, 3p.m. July 18, 3p.m. Average . 1915. 3- 4- 6 5- 7 6- 8 7- 9 8-10 9-11 6- 8 "F. 77.08 77.42 77.50 77.66 78.05 78.86 79.54 78.01 TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 15 SECOND BROOD OF NYMPHS. Length of stages. — ^The length of tlie stages and of the entire nym- phal period of this brood was shghtly shorter than in the case of first-brood nymphs. Table VIII shows that the average nymphal period of second-brood nymphs was 15.8 days; the maximum, 17 days; the minimum, 15 days. Table VIII. — Length of nymphal stages of the second brood of the a.pple leafhopper at West Chester, Pa., 1915. SECOND BROOD, JULY AND AUGUST. Aver- age mean Date of Total tem- First First Second Sec- ond stage. Third Third Fourth Fourth Fifth Fifth nym- phal pe- pera- ture for hatch- molt. stage. molt. molt. stage. molt. stage. molt. stage. ing. riod. nym- phal pe- riod. 1915. 1915. Days. 1915. Days. 1915, Days. 1915. Days. 1915. Days. Days. op July 14 July 17 3 July 19 2 July 22 3 July 25 3 July 30 5 16 79.70 July 17 July 19 2 July 23 4 July 26 3 July 30 4 Aug. 3 4 17 80.22 Do... July 20 3 July 22 2 July 24 2 July 28 4 Aug. 1 4 15 80.50 July 20 July 22 2 July 23 1 July 27 4 July 30 3 Aug. 5 6 16 79.05 Do... ...do.... 2 July 25 3 July 28 3 ...do.... 2 Aug. 4 5 15 79.00 Do... ...do.... 2 ...do.... 3 July 29 4 Aug. 1 3 Aug. 5 4 16 79.05 July 21 July 21 3 July 28 4 July 31 3 Aug. 3 3 Aug. 7 4 17 78.27 July 22 July 25 3 July 29 4 ...do.... 2 Aug. 2 2 Aug. 6 4 15 79.40 July 23 ...do.... 2 ...do.... 4 July 30 1 ...do.... 3 Aug. 7 5 15 78.56 Do... July 26 3 ...do.... 3 ...do.... 1 Aug. 3 4 Aug. 8 5 16 78.52 Do... July 25 2 July 28 3 ...do.... 2 Aug. 2 3 Aug. 7 5 15 78.56 Do... July 26 3 ...do.... 2 ...do.... 2 Aug. 3 4 Aug. 8 5 16 78.52 Do... ...do.... 3 July 27 1 ...do.... ....3 ...do.... 4 ...do... 5 16 78.52 Do... July 27 4 July 29 2 Aug. 1 3 Aug. 5 4 Aug. 9 4 17 78.38 Average 2.6 2.7 2.5 3.2 4.6 15.8 79.01 SECOND BROOD OP ADULTS. Longevity. — In Table IX the date of death for 40 second-brood adults is recorded. Table IX. — Longevity of second-brood adults; 40 adults placed in cage Aug. IS, 191,5. Date of examination. Aug. 17 Aug. 23 Aug. 26 Sept. 1. Do. Sept. 13 Do. Number dead. Date of examination. Sept. 16 Sept. 20 Sept. 24 Sept. 27 Sept. 30 Do.. Number dead. 1 2 2 3 2 •1 Escaped. Extent of reproduction. — Four separate experiments were made to determme the number of eggs deposited by a female of the second brood. Two, three, four, and five pairs of adults, respectively, were 16 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. placed in separate cages and a record of the number of eggs hatched was kept. The average number of eggs deposited per female was 19.5 in cage 1, 21 in cage 2, 20.2 in cage 3, and 18.4 in cage 4. The total average for the four cages was 19.8 eggs per female. (See Table X.) Table X. — Number of third-brood nymphs produced by second-brood adult apple leaf- hoppers in confinement. Cage No. 1.— TWO PAIRS SECOND-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE AUG. 5, 1915. Date of examination. Aug. 16. Aug. 23 . Aug. 30. Number of third- brood nymphs removed. Date of examination. Sept. 7 Average per female . Number of tliird- brood nymphs removed. 19.5 Cage No. 2.— THREE PAIRS SECOND-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE AUG. 6, 1915. Aug. 16 . Aug. 23 . Aug. 31 . Sept. 4., Sept. 14. Sept. 28. Average per female. . . 21 Cage No. 3.— FOUR PAIRS SECOND-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE AUG. 5, 1915. Aug. 16 . Aug. 23 . Aug. 31. Sept. 4.. 37 Sept. 14. 24 Sept. 28. Average per female Cage No. 4.— FIVE PAIRS SECOND-BROOD ADULTS PLACED IN CAGE ON YOUNG APPLE TREE AUG. 9, 1915. Third Generation, third brood op eggs. Length of incubation period. — Three separate cages were used in the experiments on the length of the incubation period of third- brood eggs. The average iacubation period was 9.6 to 10.3 days for cage 1 ; 9.5 to 11.5 days for cage 2; and 10 to 12 days for cage 3. (See Table XI.) TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 17 Table XJ. — Length of incubation period of third-brood eggs of the apple leafhopper. West Chester, Pa., 1915. Cage No. 1.— SIX PAIRS OF ADULTS PLACED ON APPLE TREE AUG. 6, 1915, 8 A. M.; ADULTS REMOVED AUG. 6, 4 P. M. Date of removal of newly hatched nymphs. Number of nymphs removed. Incuba- tion period. Average tempera- ture for entire period. Aug. 15, 4 p. m. Aug. 16, 4 p. m.. Aug. 17, 4 p. m.. Average.. 1915. Days. 8.6- 9.3 9.fr-10.3 10.6-11.3 'F 75.95 76.40 76.00 9.6-10.3 76.11 Cage No. 2.— TEN PAIRS OF ADULTS PLACED ON APPLE TREE AUG. 9, 1915, 4 P. M.; ADULTS REMOVED AUG. 10, 4 P. M. Aug. 19, 4 p. m. Aug. 20, 4 p. m. Aug. 21, 4 p. m. Aug. 22, 4 p. m. Average . . 1915. Days. 8-10 9-11 10-12 11-13 9.5-11.5 'F 75.15 75.00 74.87 75.07 75.02 Cage No. 3.- -TWENTY-FIVE PAIRS OF ADULTS PLACED ON APPLE TREE AUG. 13, 1915, 4 P. M.; ADULTS REMOVED AUG. 14, 3 P. M. Aug. 23, 4 p. m. Aug. 24, 4 p. m. Aug. 25, 4 p. m. Aug. 26, 4 p. m. Aug. 27, 4 p. m. Average... 1915. Days. 8-10 9-11 10-12 11-13 12-14 10-12 'F 74.65 74.95 74.70 74.57 73.89 74.55 Table XII. — Length of ntfmphal stages of the third brood of the apple leafhopper at West Chester, Pa., 1915. THIRD BROOD, AUGUST AND SEPTEMBER. Aver- age mean Date of Total tem- First First Second Sec- ond stage. Third Third Fourth Fourth Fifth Fifth nvm- phal ri(S. pera- ture for nym- phal hatch- ing. molt. stage. molt. molt. stage. molt. stage. molt. stage. ri^. 1915. 1915. Days. 1915. Days. 1915. Days. 1915. Days. 1915. Days. Days. 'F. Aug. 12 Aug. 14 2 Aug. 17 3 Aug. 20 3 Aug. 22 2 Aug. 27 5 15 74.44 Do. Aug. 16 4 Aug. 19 3 Aug. 21 2 Aug. 25 4 Aug. 30 5 18 73.10 Do.. Aug. 15 3 ...do i Aug. 22 3 ...do 3 Aug. 31 6 19 72.55 Do.. Aug. 16 4 ...do 3 Aug. 23 4 Aug. 27 4 Sept. 3 7 22 72.32 nj.*. Aug. 15 1 Aug. 18 3 Aug. 21 3 Aug. 24 3 Aug. 50 6 16 72.50 Aug. 16 2 Aug. 19 3 Aug. 22 3 Aug. 25 3 Aug. 31 6 17 71.94 Do.. ...do.... 2 ...do 3 ...do 3 ...do 3 Sept. 1 7 18 71.78 Do.. ...do.... 2 ...do 3 Aug. 23 4 Aug. 27 4 Sept. 2 6 19 71.52 Do.. Aug. 17 3 ...do 2 Aug. 2'4 5 Aug. 26 2 Sept. 1 6 18 71.78 Aug. 17 Aug. 20 3 Aug. 23 3 Aug. 26 3 Aug. 31 5 Sept. 6 6 20 71.76 Do.. ...do.... 3 Aug. 22 2 Aug. 25 3 ...do 6 ...do.... 6 20 71.76 Do.. Aug. 21 4 Aug. 23 2 ...do 2 Sept. 1 7 ...do.... 5 20 71.76 Do.. ...do.... 4 ...do 2 Aug. 26 3 ...do 6 Sept. 7 6 21 72.09 Do.. Aug. 20 3 Aug. 22 2 Aug. 25 3 ...do 7 4.2 Sept. 6 5 20 71.76 Average 2.8 2.7 3.1 5.8 18.7 72.21 132816°— IS 18 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. THIRD BROOD OF NYMPHS. Length of stages. — The average nymphal period was 18.7 days; the maximmn, 22 days; the minimum, 15 days. (See Table XII.) THIRD BROOD OF ADULTS. Attempt to rear a fourth brood of nymphs.— Third-hrood adults which were reared in confinement were placed on young apple trees in globe cages in an attempt to rear a fourth brood of nymphs. Four experiments were carried out, pairs of adults being confined on August 24, September 14 and 30, and October 1. None of the adults was observed to mate and not a single fourth-brood nymph developed. The foliage was examined several times covering a period of six weeks or more after the adults were confined. Number of winter eggs deposited. — Four globe cage experiments were conducted to determine whether the third and last brood adults of the apple leafhopper deposited winter eggs in the bark of apple trees. Pairs of adults were confined in four separate cages on August 24, September 14, September 30, and October 1. They were allowed to remain undisturbed on the trees until December 10, when the trees were brought into the laboratory and carefully examined for winter eggs under the bark. Not a single winter egg was found on any of the four trees. Table XIII. — Daily maximum, minimum, and average temperatures taken at West Chester, Pa., from May 12 to Sept. 18, 1915. Day of month. May. June. July. August. September. Maxi- mum. Mini- mum. Av- erage. Maxi- mum. Mini- mum. Av- erage. Maxi- mum. Mini- mum. Av- erage. Maxi- mum. Mini- mum. Av- erage. Maxi- mum. Mini- mum. Av- erage. 1 'F. 'F. 'F. 'F. 80 60 59 80 84 75 84 80 76 76 84 88 86 89 84 83 84 83 83 80 77 65 69 75 81 84 78 78 78 80 "F. 64 54 48 69 70 70 71 70 66 61 74 75 78 76 72 75 74 77 76 73 65 60 61 60 66 69 68 65 67 75 'F. 72 57 53.5 69.5 77 72.5 77.5 75 71 68.5 79 81.5 82 82.5 78 79 79 80 79.5 76.5 71 62.5 65 67.5 73.5 76.5 73 71.5 72.5 77.5 "F. 86 84 88 84 80 76 82 80 83 86 80 87 84 85 84 89 91 92 90 79 78 84 81 80 84 84 82 82 86 90 91 'F. 76 68 76 75 72 68 72 68 66 76 73 72 75 71 74 74 83 82 82 72 68 74 68 74 74 72 69 74 78 80 74 "F. 81 76 82 79.5 76 72 77 74 74.5 81 76.5 79.5 79.5 78 79 81.5 87 87 86 75.5 73 79 74.5 77 79 78 7.5.5 78 82 85 82.5 'F. 96 91 75 84 83 80 68 84 82 80 80 79 85 84 84 85 75 70 76 82 76 83 83 84 76 76 70 66 66 74 68 '•F. 82 79 67 76 77 70 63 72 70 70 72 71 78 73 74 77 68 58 64 65 71 72 73 72 68 70 60 62 60 68 57 'F. 89 85 71 80 80 75 65.5 78 76 75 76 75 81.5 78.5 79 81 71.5 64 70 73.5 73.5 77.5 78 78 72 73 65 64 63 71 62.5 "F. 72 73 81 83 84 86 86 86 86 90 86 84 88 88 86 88 88 81 'F. 66 66 66 72 74 76 74 71 78 76 78 74 82 80 80 76 74 70 'F. 69 2 69.5 3 73.5 4 77.5 6 79 6 7 8 9 10 11 12 13 14 15 16 17 18 19 74 77 70 63 66 68 72 80 81 79 62 71 61 74 75 66 84 72 70 76 77 79 70 69 79 80 67 65 60 54 55 58 60 59 60 54 50 62 50 52 64 59 69 66 58 68 60 64 66 63 71 70 70.5 71 65 58.5 60.5 63 66 69.5 70.5 66.5 56 66.5 55.5 63 69.5 62.5 76.5 69 64 72 68.5 71.5 68 66 7.'-. 81 80 78.5 82 83 82 79 85 84 83 82 81 75.5 20 21 22 23 24 25 26 27 28 29 30 31 TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 19 The foregoing experiments indicate that third-brood adults of the apple leafhopper, under Pennsylvania conditions, do not deposit winter eggs, but merely feed on the foliage until the time arrives for them to seek shelter for the winter. The daily temperature for the period of seasonal activity is given in Table XIII. SUMMARY OF SEASONAL HISTORY. The apple leafhopper passes the winter in the adult stage under rubbish in the nursery or more often under accumulations of leaves in adjoining woodlands. In the spring the overwintered adults make their appearance on the trees during the latter part of May and they feed on the underside of the terminal leaves for about 10 days before mating. The females deposit their eggs in the veins of the terminal leaves, the average length of the incubation period of the first-brood eggs being approximately one week. The feeding period of this brood of nymphs extends from May 30 until about the middle of July, the nymphs bemg most abundant during the third week in June. The length of the first-brood nymphal period varies from 15 to 22 days, the average being 18.7 days. First-brood adults continue to emerge from June 20 to July 20. (See fig. 2.) so Jo 20 JO :to JO OC/T ZO JO go .^9 / miff or OACC '^ , o^J. !^?\ /^^. VTMn N /o ■aOwo 9^iit\ rj /^ »J■r^ ?/ooa rV \ trOrn K^ r ^ TOOOI vooO- >v \ "^ Moon d. ^^ ■/fe*w Fig. 2.— Seasonal history of the apple leafhopper at West Chester, Pa., 1915. Second-brood eggs hatch from the latter part of June until about August 1, the length of the incubation period being about one week. The second-brood nymphs commence to appear about the last part of Jime, the larger percentage having developed into adults by the middle of August. The average length of the second nymphal period is 15.8 days; the maximum and minimum, 17 and 15 days, respect- ively. Second-brood adults appear on the trees from the middle of July until about a month later. 20 BULI.ETIN 805, TJ. S. DEPARTMENT OF AGRICULTURE. The third-brood eggs hatch from July 30 until about September 1, the length of the egg stage of this brood being about 9 to 10 days. Third-brood nymphs are found on the foliage from August 1 until about the third week in September, the average length of the entire nymphal period being 18.7 days or approximately the same as that of the first brood. By the middle of August the nymphs begin to transform to adults and by the end of September all the nymphs have disappeared. The third-brood adults remain on the trees until cold weather sets in when they gradually disperse to the hibernating places. NATURAL ENEMIES. The apple leafhopper is evidently quite free from the attack of parasites. Only one record of parasitism has appeared in the liter- ature, R. L. Webster in 1913 having noted the pupa of an egg parasite which, however, died before reaching maturity. A few cases of parasitism of adults by a dryinid have been seen at West Chester, Pa., but no adults of this parasite were reared success- fully. However, two dryinid females were captured in the field while in the act of ovipositing in the abdomen of nymphs of Empoasca mali. These specimens were determined by Mr. J. C. Crawford, of the United States National Museum, as Aphelopus alhopidus Ashm. Probably the most effective enemy of this leafhopper is the preda- cious heteropteron Triphleps insidiosus Say. This small black insect, which is rather common on the foliage during midsummer, feeds on the nymphs by thrusting its beak into their soft bodies. It is actually of little importance, however, in reducing the numbers of leafhopper nymphs. Spiders and various species of mites have been noted attacking and devouring nymphs on several occasions. In one instance an adult of the pear-leaf blister-mite was found preying on a first-stage nymph. Ladybeetle larvas also are predacious on nymphs to a small extent. R. L. Webster (18) records larvae of aphis lions and a dipteron of the family Empididae as feeding on the nymphs in Iowa. As in the case of other species of leafhoppers, both the nymphs and the adults are often caught in spider webs. THE ROSE LEAFHOPPER, Empoa rosae (Linn.). HISTORY. The rose leafhopper was originally described by Linnaeus (21, p. 439) in Europe in 1758 as Cicada rosae. Since that time this insect has been placed in several genera by various writers in Europe and North America. It was first listed by Burmeister in 1835 as TypTilo- cyha rosae Linn., and under this name it has been commonly known in Bui. 805, U. S. Dept. of Agriculture. Plate IV. s -a a 5 Bui. 805, U. S. Dept. of Agriculture. PLATE V. The Rose Leafhopper. A, Front view of head of male; B, overwiutoring eggs under bark of apple twig; C, adult with wings spread; Z), adult with larva of dryinid parasite protruding from the body; E, Anagrus armatus nigriveniris, parasite of winter egg. TWO T.EAFHOPPKrxS TN.TTTTITOTTR TO APPLE NtTRSEEY STOCK. 21 entomological literature. The genus Empoa was erected by Fitch (22, p. 63) in 1851, and in 1889 Weed (25, p. 155) transferred rosae from Typhlocyba to Empoa. In recent years rosae, with a few excep- tions, has been referred to under the genus Typhlocyba, owing to the persistant ignoring of Empoa Fitch. This insect has been known as a pest of cultivated roses in several European countries for more than a hundred years. The first account of the rose leafhopper in this country was published by Harris (23, p. 199) in 1852, when he described it as TeUigonia rosae. Al- though brief mention has been frequently made of this hopper since Harris's time, the first record of it as an enemy of apple was made by Parrott (26) in 1909. Wilson and Childs (27) in 1915 were the fu-st authors to treat of this insect at any length. They made a study of the rose leafhoppe" as a fruit pest in Oregon, giving the life history, habits, destructiveness, and remedial measures. Brittain (28), also in 1915, discussed this species as an enemy of apple in Nova Scotia. SYNONYMY. Empoa rosae (Linn.). Cicada rosae Linn. Syst. Nat., ed. 10, v. 1, 1758, p. 439. Typhlocyba rosae Burm. Handb. d. Ent., 2, 1835, p. 107. Cicadula rosae Zett. Ins. Lap., 1840, p. 299. Typhlocyba 2yteridis Dahlb. KongL Vet.-Akad. HandL, 1850, p. 179. TeUigonia rosae Harr. Ins. InJ. to Veg., 2nd ed., 1852, p. 199. Eupteryx rosae Marsh. Ent. Mo. Mag., v. 3, 1866-1867, p. 246. Typhlocyba lactea DougL Ent. Mo. Mag., v. 12, 1875, p. 77. Anomia rosae Fieber. Rev. d'Ent., v. 3, 1884, p. 124. Empoa rosae Weed. Amer. Gard., July, 1889, p. 257. ORIGIN AND DISTRIBUTION. The rose leafhopper is undoubtedly of European origin, as it was known to Linnaeus more than a century before it was first recorded m this country. It probably was introduced from abroad in the egg stage on rose or apple stock. It is distributed generally throughout the United States and has been taken from States forming the extreme northern, southern, eastern, and western limits of the country. It is apparently most abundant in the Northern States, particularly in the Pacific North- west. In Canada, this insect has been reported from Nova Scotia, Ontario Province, and from British Columbia. In Europe it has been recorded from several localities in England, France, and Germany. FOOD PLANTS. This insect, although j^rimarily a pest of rose and apple, is a rather general feeder. At West Chester, Pa., it has been taken feeding upon rose, apple, pear, peach, plum, cherry, quince, currant, goose- berry, raspberry, blackberry, grape, Crataegus, elm, oak, and cotton- 22 BULLETIN 805, V. S. DEPARTMENT OF AGRICULTURE. wood (Platanus deltoides). In addition, Wilson and Childs list strawberry, logan blackberry, and prune from Oregon. It is likely that this leafhopper will feed on the foliage of most plants belonging to the family Bosaceae. CHARACTER OF INJURY. The nymphs and adults of Empoa rosae confine their feeding entirely to the lower leaves of apple trees in the nursery. They congregate on the lower surfaces of the foliage and suck the plant juices by puncturing the leaf tissue with their tiny beaks. The first indication of the injury is the mottling of the leaves with yellowish or whitish spots at the points where the pimctures were made. (See PI. IV.) When the leaves become heavily infested they turn yellow, dry up, and drop to the ground prematurely. The foliage is never curled by this species, nor is the terminal growth checked, as in the case of injury by Empoasca mali. A second type of injury is produced by the egg pimctures made by the females in the fall during the oviposition period. The eggs are deposited under the bark of young apple trees, several hundred eggs often being placed in a single twig. The injury produced by the rose leafhopper to apple nursery stock is of little importance, however, when compared with that caused by the more destructive apple leafhopper. DESCRIPTION OF STAGES. PI. V, B. The winter egg is elongate oval, slightly crescentic xn form, almost circular in cross section, and blunt at both ends. It is almost transparent at first, but when ready to hatch it changes to a milky white color, while the red eyes of the young nymph are visible through the smooth chorion. Average length of 16 eggs 0.65 mm., width 0.19 mm. First instar. — Color of first-instar nymph pale white changing to light yellow after feeding. Eyes dull red. Small spines present on the dorsal side of the head, thorax, and abdomen; the latter with four spines to each segment arranged in two longitu- dinal rows on each side. Posterior margin of metathorax blunt. First two segments of antennse pale, remainder dusky. Average length of 16 specimens 0.98 mm. Second instar. — General color creamy white to light yellow. Eyes lose some of their red color, becoming lighter. Wing pads begin to appear as lateral buds. Pos- terior margin of metathorax sharp in outline. First two segments of antennae yellow, remainder dusky. Average length of 16 specimens 1.27 mm. Third instar. — General col or light yellow. Eyes dull white. Body more robust than in first two stages. Wing pads extending to hind margin of the first abdominal seg- ment. Spines darker and more prominent. Average length of 16 specimens 1.55 mm. TWO LEAFHOPPERS IJIJURiOUS TO APPLE NURSERY STOCK. 23 Fourth instar. — General color light yellow. Eyes almost pearl white with a brown central spot underneath. Wing pads extending to hind margin of the second abdom- inal segment. Spines very distinct. Average length of 16 specimens 2.03 mm. Fifth instar. — General color as in previous stage. Eyes almost pearl white. Wing pads extending nearly to the hind margin of the fourth abdominal segment. Broader than in previous stage. Average length of 16 specimens 2.85 mm. THE ADULT. n. V, A, c. General color of adult creamy white to light yellow. Head rather pointed, pale yellow; face of male with a tint of orange color in the form of a central longitudinal stripe with several transverse radiating stripes; gense pale, narrow, and almost as long as clypeus; lorse very narrow, sunken beneath the compound eyes. Vertex pale yellow with two eemitransparent spots just above the ocelli, a faint median longitudinal semitransparent line extending from the hind margin half-way to the front margin, and a semitransparent line bordering each eye. Two ocelli present, situated on the frontal margin of the vertex, marked by two white spots with a dark center, distance apart twice that from the eye to the ocellus; eyes pearl white with a da,rkened center, ashen-gray after death. Pronotum light yellow with a semitransparent area in the center; mesonotum semitransparent with a large trapezoidal creamy area extending from the front to the hind margins, wider behind; a small cream colored area anteriorly on each side; scutelliun cream-colored, sometimes with a semitransparent area caudad. Elytra transparent. First two segments of antennae pale, flagellum dusky. Legs pale, tarsi dusky at tips. Sexual appendages slightly ciliated in female only. Average length of 12 specimens 3.00 mm. life history and habits. The Egg. Part of the rearing work in the life history of the rose leafhopper was conducted during the spring of 1915, the work being completed during the season of 1916. In obtainuig data on the biology of this insect the same methods as devised for the study of the apple leaf- hopper were used. V/inter egg. — -This leafhopper hibernates in the egg stage, winter eggs bemg deposited in the fall during the period extending from the last week in September to November 1. They are laid almost entirely under the bark of apple trees and rose bushes, though a few have been found in pear, quince, cherry, plmn, currant, and Cra- taegus. The eggs are deposited singly, usually lying in a longitu- dinal position just under the epidermis, and they are placed in a dis- tmct blister or pouch which measures 0.7 to 0.8 mm. m length. They are easily located by looking on the bark for the raised blister, which, as a rule, is slightly crescentio in outline. Winter eggs have been found in the bark of nursery apple trees from one to four years old; they are found in greatest numbers on the first year's growth of two-year trees. The favorite location for oviposition seems to be around the bases of the lowest limbs or on the trunk just below the first branches. 24 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. Summer c^grs,— Summer eggs are laid by first-brood adults during July in the veins of the lower leaves. The Nymph. The n}Tnphs of this species diHer from, those of Empoasca maJi by their paler color, by their smaller size, and by the fact that they confine their activities entirely to the lower leaAes of the trees. Though very active they are not quite as quick in their movements as the nymphs of the apple leafhopper. As the yomig nymphs emerge from the eggs in the bark they make their way to the nearest leaves, where they immediately settle down to feed. In the vicinity of West Chester, Pa., the first nymphs of the season emerge about May 1, and by May 15 practically all the winter eggs have hatched. Generally speaking, the nymi^hs of this species are from three to four weelcs old before the first nymphs of Empoasca mail appear on the terminal leaves. The Adxtlt. Nymphs of the first brood transform to adults dm*ing the first two weeks in June and these adults feed on the foliage for several weeks before mating. Oviposition extends over a period of about two weeks during late June and early July, most of the first-brood adults dying by the end of July. Second-brood adults begm to appear during the first week in August, and they remain on the trees until death, which occurs by Novem- ber 1, after the winter eggs have been deposited. A few adults have been noticed on the trees as late as November 25. No rose leafhoppers, either in confinement or in the field, were observed to hibernate in the adult stage. First Generation, first brood of eggs. The length of the incubation period of the first-brood eggs (i. e., the winter eggs) is 6 to 7 months. Eggs deposited during October hatch by the middle of May of the following year at the latest. FIRST BROOD OP NYMPHS. Nymphs newly hatched from winter eggs were confined in indi- vidual cork cages on uninfested leaves. The average length of the entire nymphal period, as indicated in Table XIV, was 33.4 da^^s; the maximum, 36 days; the minimum, 30 days. TWO LEAFHOPPEES INJURIOUS TO APPLE NURSERY STOCK. 25 Table XIV. — Length ofnymphal stages of the first brood of the rose leaf hopper on nursery apple trees at West Chester, Pa., 1915. First brood. May and June. Date of hatch- ing. First molt. First stage. Second molt. Sec- ond stage. Third molt. Third stage. Fourth molt. Fourth stage. Fifth molt. Fifth stage. Total nym- phal pe- riod. Aver- age mean tem- pera- ture for nym- ph al pe- riod. 1915. May 6 May 7 Do... Do... Do... Do... Do... Do... May 8 May 11 Do... 1915. May 11 May 13 May 14 ...do.... May 15 May 12 May 16 May 15 May 13 May 16 May 17 Days. 5 6 7 7 8 5 9 8 5 5 6 1915. May 14 May 17 May 21 May 22 ...do May 20 May 23 ...do May 21 ...do May 22 Days. 3 4 7 8 7 8 7 8 8 5 5 1915. May 21 May 24 May 29 May 28 ...do May 26 May 29 May 30 May 27 ...do May 28 Days. 7 7 8 6 6 6 6 7 6 6 6 1915. May 28 May 30 June 4 Juno 2 June 3 ...do June 5 June 4 June 3 ...do ...do Days. 7 6 6 5 6 8 7 5 7 7 6 1915. Juno 7 June 8 June 12 Juno 10 Juno 11 June 9 Juno 12 Juno 11 ...do... June 10 June 11 Days. 10 9 8 8 8 6 7 7 8 7 8 Days. 32 32 36 34 35 33 36 35 34 30 31 "F. 67.19 67.33 68.16 67.47 67.79 67.44 68.16 67.79 67.70 67.85 68.29 Average 0.4 6.3 6.4 6.3 7.8 33.4 67.74 FIRST BROOD OF ADULTS. No attempt was made to obtain definite data on the longevity of first-brood adults. Ninety adults, however, were confined in a Riley cage on a yomig apple tree on May 31, 1916, and the last leafhoppers died on August 1 ; by this date many of the second-brood n3anphs had appeared. Thus, the length of life of the adults of this brood is approximately two months. Second Generation. second brood of eggs. Length of incubation period. — ^An approximation of the length of the egg stage of second-brood eggs (i. e., the summer eggs) is shown in Table XV. The average length of the period was 25 days; the maximum, 27 days; the minimum, 23 days. Table XV. — Length of incubation period of summer eggs deposited by first-brood females of the rose leaf hopper, West Chester, Pa., 1916. Four pairs of adults placed in confinement on apple tree July 28, 1916, 4 p.m.; adults removed July 29, 4 p. m. Date of removal of newly hatched nymphs. 1916 Aug. 21 Aug. 22 Aug. 23 Aug. 24 Number of nymphs removed Incuba- tion period. Days. 2.5-25 24-26 25-27 Date of removal of newly hatched uymphs. 1916 Aug. 25 Aug. 26 Aug. 27 Aug. 28 Number of nymphs removed. Incuba- tion period. Days. 26 BULLETIN 805^ V. S. DEPARTMENT OF AGEICULTUEE. SECOND BROOD OP NYMPHS. Length of stages. — ^In Table XVI the length of the stages and of the entire nymphal period is recorded. The average length of the njmphal period was 17.7 days; the maximimi, 19 days; the mini- mum, 16 days. It will be noted that the length of the nymphal peiiod of the second brood was much shorter and that the corre- sponding temperatures were much higher than in the case of the first-brood nymphs. Table XVI. — Length of nymphal stages of second brood of the rose leaf hopper on nursery apple trees at West Chester, Pa., 1916. Second brood, July and August. Aver- age mean Date of Total tem- First First Second Sec- ond stage. Third Third Fourth Fourth Fifth Fifth nym- phal pe- riod. pera- hatch- ing. molt. stage. molt. molt. stage. molt. stage. molt. stage. ture for nvm- piial pe- riod. 1916. 1916. Days. 1916. Days. 1916. Days. 1916. Days. 1916. Days. Days. op July 21 July 25 4 July 28 3 July 31 3 Aug. 3 3 Aug. 7 4 17 79.72 July 24 July 28 4 July 31 3 Aug. 4 4 Aug. 6 2 Aug. 10 4 17 80.13 July 25 July 27 2 ...do 4 Aug. 3 3 ...do 3 ...do.... 4 16 80.38 Do... July 28 3 Aug. 1 4 Aug. 5 4 Aug. 7 2 Aug. 12 5 18 80.55 Do... July 31 6 Aug. 3 3 ...do 2 Aug. 8 3 ...do.... 4 18 80.55 July 28 ...do.... 3 Aug. 4 4 Aug. 6 2 Aug. 9 3 Aug. 15 6 18 79.84 Do... Aug. 1 4 ...do 3 ...do 2 ...do 3 ...do.... 6 18 79.84 Do... ...do.... 4 ...do 3 ...do 2 Aug. 10 4 Aug. 16 6 19 79.70 Do... ...do.... 4 Aug. 5 4 Aug. 8 3 Aug. 12 4 Aug. 15 3 18 79.84 July 30 Aug. 2 3 Aug. 6 4 ...do 2 ...do 4 Aug. 17 5 18 80.26 Do... Aug. 4 5 ...do 2 Aug. 10 4 Aug. 15 5 Aug. 18 3 19 80.40 Do... Aug. 2 3 ...do 4 Aug. 9 3 Aug. 13 4 ...do.... 5 19 80.40 July 31 Aug. 5 5 Aug. 9 4 Aug. 12 3 Aug. 16 4 Aug. 19 3 19 80.75 Do... ...do.... 5 Aug. 8 3 ...do 4 Aug. 15 3 ...do.... 4 19 80.75 Aug. 3 Aug. 6 3 Aug. 10 4 Aug. 13 3 Aug. 16 3 Au^. 20 4 17 80.94 Aug. 4 Aug. 7 3 Aug. 9 2 ...do 4 Aug. 17 4 Aug. 21 4 17 81.38 Do... Aug. 6 2 Aug. 11 5 Aug. 15 4 Aug. 18 3 ...do.... 3 17 81.38 Do... Aug. 8 4 Aug. 12 4 ...do 3 Aug. 17 2 Aug. 20 3 16 81.23 Do... ...do.... 4 Aug. 13 5 Aug. 16 3 Aug. 19 3 Aug. 22 3 18 81.68 Aug. 5 ...do 3 Aug. 11 3 Aug. 15 4 Aug. 17 2 3.2 Aug. 21 4 16 81.29 Average 3.7 3.5 3.1 4.1 17.7 80.50 SECOND BROOD OP ADULTS. Extent of reproduction. — Pairs of second-brood adults were con- fined in arc-light globe cages to determine the number of winter eggs deposited per female in the bark of apple. The data in Table XVII were obtained in the fall of 1915, and that in Table XVIII in 1916. The hoppers were allowed to remain on the trees until death. The average number of eggs deposited per female during the two experi- ments was 15.5 and 16, respectively. Attempt to rear a third hrood of nymphs. — A globe-cage experiment was conducted to determine whether second-brood adults would produce a third brood of nymphs. Ten pairs were confined on August 7 in two cages and examined at intervals until October 1 . In no case TWO LEAFHOPPEES INJURIOUS TO APPLE NURSERY STOCK. 27 did any third-brood nymphs develop, only winter eggs being depos- ited by the females. Table XVII. — Number of winter eggs deposited in bark of apple trees by second-brood adults in confinement, West Chester, Pa., 1915. Cage. Date of confinement. Number of pairs confined. Date of examination. Number of eggs present. 1 1915. Sept. 26 6 5 1915. Nov. 25 98 2 Oct. 9 .. ...do 78 Total 176 16 Table XVIII. — Number of winter eggs deposited in bark of apple trees by second-brood adults in confinement, West Chester, Pa., 1916. Cage. Date of confinement. Number of pairs confined. Date of examination. Number of eggs present. 1 1916. 6 9 1916. Nov. 7 104 2 Sept 6 Nov. 8 lae Total 233 15.5 In Table XIX tbe daily temperature is given for the period of seasonal activity, 1916. Table XIX. — Daily maximum, minimum, and average temperature taken at West Chester, Pa., from July SI to Oct. 81, 1916. July. August. September. Octoljer. Day of month. Maxi- mum. Mini- mum. Aver- age. Maxi- mum. Mini- mum. Aver- age. Maxi- muiu. Mini- mum. Aver- age. Maxi- mum. Mini- mum. Aver- age. 1916. 1 "F. "F. "F. °F. 85 80 78 88 88 92 91 92 85 82 86 86 74 78 80 80 90 86 90 88 93 92 92 79 84 84 86 70 74 80 82 op 74 70 74 78 77 80 82 75 76 68 78 78 70 72 74 74 80 80 82 82 75 82 82 69 66 75 74 64 64 66 65 °F. 79.5 75 76 83 82.5 86 86.5 83.5 80.5 75 82 82 72 75 77 77 85 83 86 85 84 87 87 74 75 79.5 80 67 69 73 73.5 °F. 84 81 72 78 83 84 86 88 76 74 72 76 81 84 72 66 70 68 62 72 72 75 76 68 66 72 79 82 72 59 "F. 78 76 66 70 72 76 78 74 68 66 60 61 65 72 64 56 62 60 51 57 56 64 68 55 57 57 66 70 56 46 "F. 81 78.5 69 74 77.5 80 82 81 72 70 66 68.5 73 78 68 61 66 64 56.5 64.5 64 69.5 72 61.5 61.5 64.5 72.5 76 64 52.5 °F. 53 65 70 65 75 73 70 77 82 53 62 70 56 60 60 62 68 58 67 80 57 54 56 61 64 56 58 56 61 67 62 °F. 45 50 53 61 59 65 62 60 70 43 50 58 54 52 45 59 60 44 56 68 49 45 44 48 46 48 49 49 46 50 50 "F. 49 2 57.5 3 61.5 4 63 5 67 6 69 7 66 8 68.5 9 76 10 48 11 56 12 64 13 55 14 56 15 52.5 16 60.5 17 64 18 51 19 61.5 20 74 21 80 81 84 78 76 82 90 80 80 84 92 72 70 76 74 74 78 75 74 74 74 85 76 75.5 80 76 75 80 82.5 77 77 79 88.5 53 22 . . 49.5 23 50 24 54.5 25 55 26 . . 52 27 53.5 28 52.5 29 53.5 30 58.5 31 56 28 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. SUMMARY OF SEASONAL HISTORY. There are two generations of the rose leafhopper annually at West Chester, Pa. This insect hibernates in the egg stage, the eggs being deposited under the bark of the host plants. On apple stock in the nursery the winter eggs hatch from May 1 to May 15, and the newly hatched nymphs immediately attack the lower leaves of the trees. The feeding period of the first brood of nymphs covers approximately one month. The first adults of the season appear by the end of May. The latter feed for several weeks before mating and depositing the second brood of eggs in the foliage. The length of the life of the first-brood adults is about two months, most of them having dis- appeared by the first week in August. The length of the incubation period of the second-brood eggs is about 25 days, the first eggs hatching about July 20. Due to the higher temperatures, the length of the nymphal period of this brood is comparatively shorter than that of the spring brood of nymphs. Second-brood nymphs, on an average, attain the adult stage in 17 days. By the latter part of August, practically all the nymphs have transformed to adults. The adults feed for about a month before mating, which takes place during the latter part of September. Females deposit the winter eggs throughout the month of October, soon after which they die. The last adults of the season were found on the trees on November 25. NATURAL ENEMIES. The most efficient enemies of the rose leafhopper are two species of hymenopterous egg parasites belonging to the superfamily Proc- totrypoidea. Thesepar asites were determined by Mr. A. A. Girault, of the Bureau of Entomology, as Anagrus epos Girault and Anagrus arT/iatus Ashm.. var. nigriventris Girault. (See PL V, E.) These two parasites were reared from the winter eggs only, and they emerged both in the fall and in the spring. In the fall they emerged from October 15 to November 1, while in the spring they were reared in abundance about one or two weeks after the hatching of the winter eggs. These egg parasites are a valuable factor in reducing the destructive numbers of the leafhopper. At West Chester, Pa., from 65 to 70 per cent of the winter eggs were parasitized during 1916. Both of these species evidently are widely distributed, as they have been reared from the eggs in apple twigs from Hagerstown, Md., Winchester, Va., and Roswell, N. Mex. Adults of this leafhopper are parasitized quite heavily by a species of the family Dryinidae. A rose leafhopper parasitized by a dryinid (PI. V, D) is readily recognized by the distorted appearance of one of its wings under which the ])arasitic larva is noticed protruding TWO LEAFHOPPEES INJURIOUS TO APPLE NURSERY STOCK. 29 from the abdomen. Numerous parasitized leafhoppers were collected but no adult parasites were reared successfully. A. Giard (24), in 1889, reported the dryinid Aphelopus melaleucus Dalm. and the pipunculid Ateleneura spuria Meig. as parasites of Ty2)Mocyha rosae in France. Among the predacious enemies of this insect, mites, spiders, and coccinellid and syrphid larvsB have been noted feeding on the nymphs to a limited extent. Wilson andChilds (27), in 1915, recorded the larva of a green lace- wing, Chrysopa sp., as preying on young nymphs in Oregon. The same authors list a dragonfly and a scatophagidfly as predacious en,emies of adults. REMEDIAL MEASURES. The apple leafhopper, Empoasca mali, is one of the most difficult of all leafhoppers to combat successfully, especially because of the wide range of its food plants. In nurseries, however, this species con- fines its attack chiefly to the foliage of the apple, and if the hoppers can be reduced in sufficient numbers on this plant the danger of infesta- tion from other host plants will be very small. Little actual work has been done in dealing with the remedial measures to be used against this pest. One of the most frequently recommended methods which many writers have suggested as a means of controlling this leafhopper in nurseries is the use of sticky shields to catch the winged adults as they are jarred from the trees along the nursery row. The use of such a device has never proved satisfactory and it is impracticable when employed on a large scale. Spraying with contact insecticides against the apple leafhopper has been generally advised. Weiss (19) suggested spraying with 1 pint of "black-leaf 40" to 100 gallons of Pyrox. But in general the recommendations have been so indefinite and vague that little value can be attached to them. In order to clear up the confusion which has existed in regard to the treatment for this insect, spraying experi- ments, based on the life-history studies, were carried out. SPRAYING FOR THE FmST BROOD. Spraying ex:periments against the nymphs were conducted in the nurseiy of Hoopes Bros, & Thomas Co. at West Chester, Pa. Tests on a small scale were made, the spray being applied with a compressed- air sprayer and an angle nozzle at a pressure of 60 pounds. Larger plats were sprayed with a machine owned by the nursery which consisted of a double-acting hand pump capable of 100 pounds pres- sure and a 50-gallon galvanized tank set on a truck narrow enough to allow it to pass between two rows of trees. "Set nozzles" on an arrangement of pipes at the rear of this outfit were used, the nozzles 30 BULLETIN 805^ U. S. DEPARTMENT OF AGRICULTURE. being so placed that it was possible to spray the undersides of the leaves of two rows of trees from both sides. It was necessary to use eight nozzles on this machine and consequently the pressure was greatly reduced. Table XX indicates the results obtained with the compressed-air sprayer against the first-brood nymphs: Table XX. — -Spraying experiments against the Jirst-hrood nymphs of the apple leaf- hopper, West Chester, Pa., 1916. Plat No. Spray material. 40 per cent uicotin sulphate and fish-oil soap 40 per cent nicotin sulphate and fish-oil soap 40 per cent nicotin sulphate 40 per cent nicotin sulphate lime-sulphur solution. Check; unsprayed Strength. Date of applica- tion. Number of trees treated. 1-1,000 2-50 1-1,400 2-50 1-1,400 1-1,500 14-50 June 12 ...do.... ..do.... ..do... 300 300 300 300 Nymphs killed. Per cent. 99 It will be noted that good results were obtained on all plats of this experiment from the use of 40 per cent nicotin sulphate, though slightly better control was obtained when soap was added as a spreader. The hand-power outfit with "set nozzles" was used in an experi- ment on several thousand 2-year-old trees which were heavily in- fested with first-brood nymphs. The results are shown in Table XXIV. Table XXI. — Spraying experiments against the first-hroo d nymphs of the apple leaf- hopper, West Chester, Pa., 1916. Plat No. Spray material. Strength. Date of applica- tion. Number trees treated. Nymphs killed. 40 per cent nicotin sulphate. do do 40 per cent nicotin sulphate and flsh-oil soap. Fish-oil soap Kerosene emulsion Check, unsprayed 1-1,400 1-1,500 1-1, 60O / 1-1,500 \ 2-50 1-8 5 per cent Jmie 15 ...do ...do }june 16 .do.... .do.... 4,000 1(),(K)0 4,000 13,000 4,000 4,000 Per cent. 92 91 89 97 50 45 0 As will be noted, the best results in Table XXI were obtained on plat 4 where 40 per cent nicotin sulphate at a strength of 1-1,500 was combined with soap. The nicotin solution when used alone proved a little less effective than when the soap was added to the mixture as shown in plats 1, 2, and 3. A potash fish-oil soap solu- tion at a strength of 1 pound of soap to 8 gallons of water (plat 5) gave poor results and was but little more valuable than a 5 per cent kerosene emulsion which was tried on j^lat 6. TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 31 During the spray application with "set nozzles" it was noted that the terminals of an occasional tree were missed by the spray. One reason for this was that the machine was top heavy and easily jolted when passing over a rough or stony place, and this resulted in the spray material being sent over the top of a tree. Furthermore, a few trees would outgrow the remainder of the trees in a row, so the spray material from the ''set nozzles" would not reach the terminals of such trees. Nevertheless, sufficiently good results were obtained with the "set nozzles" to warrant their use. TREATMENT FOR THE SECOND BROOD. An attempt was made to destroy the second-brood nymphs by spraying and dipping upon trees which had received no treatment against the first brood. On the date of this experiment, July 18, the terminal leaves were so badly curled that the nymphs were well pro- tected from the action of the spray. Four plats of about 700 trees were sprayed with a compressed-air sprayer, using 40 per cent nicotin sulphate at various strengths from 1-800 to 1-1,500 with soap added. The counts showed that only 2 per cent of the nymphs were killed on the best plats. The dipping work was done with large shallow pans specially con- structed for dipping leafhoppers and aphids on young nursery trees. The tops of the trees were thorouglily immersed in the pans of spray material. The same number of plats and the same insecticides that were used in the spray treatment above were tried in the dipping experiments. The results obtained by dipping were about the same as the spraying results, less than 2 per cent of the nymphs being killed. Since such poor results were secured from spraying experiments against the second brood no effort was made to spray the third-brood nymphs, for at the time of the presence of the latter on unsprayed trees the terminals are even further curled and inaccessible to a spray liquid. METHODS AND TIME OF APPLICATION. Attempts to control the winged adults of leafhoppers by spraying have always proved futile because of the agility of movement of the full-grown insects and because of their immunity to the action of a spray liquid at this stage. This has been demonstrated by previous investigators in the case of the apple leafhopper as well as in the case of other leafhoppers of economic importance. A spray treatment to be effective must be applied when the insects are in the nymphal stages. The proper time to spray against the apple leafhopper is at the time when a large number of the first-brood nymphs have attained the third stage of development. For the 32 BULLETIN 805, V. S. DEPARTMENT OF AGRICULTURE. seasons of 1915 and 1916 at West Chester, Pa., this date was between June 10 and June 20. It is not safe to delay the apphcation untU the end of the nymphal period, since by that time the insects will have produced sufficient injury to cause the tender terminal leaves to curl, and the result is that the nymphs on the curled leaves will be protected against the spray. The nymphs feed on the undersides of the terminal leaves during the entire nymphal period, so the spray material must be directed in such a manner as to wet the underside of the leaves thoroughly. Special care must be taken to see that the foliage of the upper half of the trees and especially the terminals are well sprayed. If a machine with "set nozzles" is used, the lower nozzles should be placed at an angle that will insure a thorough spraying of the lower surfaces of the leaves. TREATMENT FOR THE ROSE LEAFHOPPER. The injury caused by the rose leafhopper on nursery apple trees is seldom serious enough to warrant a special spray apphcation. This species confines its attack to the undersides of the leaves of the lower half of the trees and the mjury is characterized by mottled white or yellowish spots on the foliage. As a result of continued feeding some defoliation takes place which reduces the vitality of the trees slightly. Whenever necessary, this species can be controlled by one spray application of a tobacco insecticide against the first-brood nymphs. A spraying of J pint of 40 per cent nicotin sulphate to each 50 gallons of water with the addition of 2 pounds of soap, made at the time when the maximum number of nymphs are present on the foliage, wUl give satisfactory results. The most effective time to make this application is when the greatest number of njTuphs have reached the third stage, which is three to four weeks earlier than the date for the spraying against the first-brood nymphs of the apple leafhopper. For southeastern Pennsylvania the correct time for spraying the rose leafhopper during the seasons of 1915 and 1916 was from May 15 to May 25. RECOMMENDATIONS. The apple leafhopper can be controlled by spraying agamst the first-brood nymphs with a tobacco insecticide. A 40-per cent nicotin sulphate solution at the rate of 1-1,500 to which 2 pounds of soap is added to each 50 gallons will so reduce their numbers that injury to the growth of the trees by the later broods will not be serious. When it is desirable to use lime-sulphur solution with the tobacco insecticide, soap must be omitted to prevent burning. The spray shoidd be directed upward so as to wet the underside of the leaves, and par- TWO LEAFHOPPERS I2!TJURI0US TO APPLE NURSERY STOCK. 33 ticular attention should be paid to wetting every terminal leaf thor- oughly. The application should be made when the majority of the nymphs are in the third stage, which occurs about three weeks after the first nymphs are found in the terminal leaves. The rose leafhopper is far less injurious to the foliage of nursery apple stock than the apple leafhopper. Should the infestation by rose leafhoppers be heavy enough to justify spraying, the same remedial treatment recommended for the apple leafhopper can be used, except that the application should be made three or four weeks earlier than the one for the latter insect. SUMMARY. The apple leafhopper, Empoasca mali, causes serious injury to apple nursery stock by extracting the plant juices from the terminal leaves; as a consequence the leaves curl, become undersized, and fail to function normally, thereby retarding the growth of the trees. The injury is produced by the feeding of both nymphs and adults. In southern Pennsylvania this species is three-brooded and hiber- nates only in the adult stage. Eggs are laid within the leaf tissue on the underside of the leaves. This leafhopper is widely distributed over the United States and attacks a great variety of host plants. In literature the above species has been confused with another leafhopper which attacks the foliage of nursery apple trees, namely, the rose leafhopper, Empoa rosae. The latter insect is two-brooded and winter is passed in the egg stage. Winter eggs are deposited under the bark of apple trees. These eggs hatch about a month earlier in the spring than eggs deposited by overwintered females of the former species. The rose leafhopper may be distinguished from the apple leaf- hopper by its lighter color and by the absence of the six or eight white spots present on the frontal margin of the pronotum of the latter species. Differentiation between the nymphs of the two species is more difficult. The distinct types of injury produced by the two insects, however, is a ready means of distinguishing them. The rose leafhopper feeds on the lower leaves and produces white or yehow spots on them while the other species attacks the terminal leaves, curls them, and stunts the growth of the trees. Parasites seem to play a far more important role in reducing the numbers of the rose leafhopper than they do in reducing the numbers of the more injurious apple leafhopper. Larvae of dryinid parasites are quite common on the adults of the former while only rarely have they been found attacking the latter. Anagrus epos Girault and Anagrusarmatus Ashjn.Y&r.nigrivenirisGiraiult, parasites of the win- ter egg of the rose leafhopper, help considerably in checking the num- 34 BULLETIN 805, U. S. DEPARTMENT OF AGRICULTURE. bers of this species. No parasites have been reared from eggs of the apple leafhopper. A single spraying with 40 per cent nicotin sulphate at the rate of 1-1,500 combined with soap will so materially check an infestation by the apple leafliopper when applied against the first-brood nymphs that injury caused later by those that escape will be of little conse- quence. The same treatment made three or four weeks earlier is effective against the rose leafhopper, though this species is seldom injurious enough to justify a special application. LITERATURE CITED. THE APPLE LEAFHOPPER. (1) Le Baron, W. 1853. Observations upon two species of insects injurious to fruit trees. In The Prairie Farmer, Sept., p. 330-331. (2) Walsh, B. D. 1862. Entomological note. In The Prairie Farmer, Sept., p. 148-149. (3) Berg, C. 1879. Hemiptera Argentina. Enumeravit Speciesque Novas, p. 273. (4) Forbes, S. A. 1883. The green apple leafhopper. In 13th Rept. State Ent. 111., p. 181-183. (5) 1886. Proceedings of the Entomological Club of the A. A. A. S. In Ent. Amer., v. 2, p. 173-174. (6) WOODWORTH, C. "W. 1889. North American Typhlocybini. In Psyche, v. 5, p. 211-214. (7) Gillette, C. P. 1890. A few injurious insects and their remedies. In Trans. Iowa Hort. Soc, V. 25, p. 102-106. (8) OSBORN, H. 1896. A new pest of potatoes. In Iowa Agr. Coll. Exp. Sta. Bui. 33, p. 603-605. (9) Gillette, C. P. 1898. American leafhoppers of the subfamily Typhlocybinae. In Proc. U. S. Nat. Mus., V. 20, no. 1138. (10) Forbes, S. A. 1900. The leafhoppers. In 111. Agr. Exp. Sta. Bui. 60, p. 410-427. (11) Washburn, F. L. 1903. A destructive leafhopper. In Eighth Rept. State Ent. Minn., p. 148- 151. (12) Britton, W. E. 1905. The apple or currant leafhopper, Empoasca mali Le B. In Fourth Rept. Sta. Ent. Conn., 1904, p. 216. (13) Brues, C. 1905. Report of the State Nursery Inspection. In 22d Ann. Rept. Agr. Exp. Sta. Univ. Wis., p. 322-329. (14) Webster, R. L. 1908. The eggs of Empoasca mali Le B. In Jour. Econ. Ent., v. 1, p. 326-327. (15) Washburn, F. L. 1908. Two years' work with the apple leafhopper. In Minn. Agr. Exp. Sta. Bui. 112, p. 145-164. TWO LEAFHOPPERS INJURIOUS TO APPLE NURSERY STOCK. 35 (16) Garman, H. 1908. Spraying apple trees. Kentucky Agr. Exp. Sta. Bui. 133. (17) Gibson, A. 1909. Insects of the year 1908 at Ottawa. In 39tli Rept. Ont. Ent. Soc, p. 116-120. (18) Webster, R. L. 1910. The apple leafhopper. Iowa Agr. Exp. Sta. Bui. 111. (19) Weiss, H. B. 1912. Insect Record. In New Jersey Agr. Exp. Sta. Rept., p. 423-450. (20) Webster, R. L. 1915. Potato insects. Iowa Agr. Exp. Sta. Bui. 155. THE ROSE LEAFHOPPEK. (21) Linnaeus, 0. 1758. Systema Naturae. Ed. 10, v. 1. (22) Fitch, A. 1851. Catalogue with references and descriptions of the (homopterous) insects collected and arranged for the State cabinet of natural history, New York. In Fourth Ann. Rept. Sta. Cab. Nat. Hist., Albany, p. 43-69. (23) Harris, T. W. 1852. Insects injurious to vegetation. Ed. 2. (24) GiARD, A. 1889. Sur une galle produite chez le Typhlocyba rosae L. par une larve d'Hymenoptere. (p. 79-82.) Sur la castration parasitaire des Typhlocyba. (p. 708-710). Jm Comptes Rendus Ac. Sc. Paris, v. 109. (25) Weed, C. M. 1889. Notes on some little known injurious insects. In Ohio Agr. Exp. Sta. Bui., V. 6, no. 2, 2nd ser. (26) Parrott, p. J. 1909. Scientific notes. /«. Jour. Econ. Ent., v. 2, p. 79. (27) Wilson, H. F., and Childs, L. 1915. The rose leafhopper as a fruit pest. In 2nd Bien. Crop Pest Rept., Ore. Agr. Coll. Exp. Sta. 1913-1914, p. 189-194. (28) Brittain, W. H. 1915. Some Hemiptera attacking the apple. In Proc. Ent. Soc. Nova Scotia, August 3, p. 7-47. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRrNTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 807 Contribution from the Bureau of Entomology L. O. HOWARD, Chief J^^'^^U Washington, D. C. PROFESSIONAL PAPER January 27, 1920 THE BROAD-BEAN WEEVIL. By Roy E. Campbexl, Scientific Assistant, Truclc Croi) Insect Investigations. CONTENTS. Page. Introduction 1 Description 2 Synonymy 3 Records of occurrence in California- 3 Distribution in California 4 Spread in California 4 Dissemination 4 Nature of damage 5 Extent of damage 6 Life history 8 Seasonal history 11 Page. Germination tests of infested seeds- 13 Natural enemies 14 Control measures 15 Dry heat .- 15 Fumigation 16 Holding over seed 17 Late planting 18 Recommendations 21 Summary 21 Literature cited 22 INTRODUCTION. The growing of broad or horse beans in California during the last few years has been seriously handicapped everywhere by the presence of the broad-bean weevil {Bruchus rufimamis Boh.), The practi- cal impossibilit}' of growing uninfested beans caused the early aban- donment of a considerable acreage. The greatest abandonment, how- ever, followed the ruling, according to the Federal Food and Drugs Act, that weevil-infested broad beans are adulterated food, and that their shipment was prohibited in interstate commerce. The Bureau of Chemistry first ruled that beans infested " to any material extent " could not be shipped (4),^ but later changed the rul- ing to allow shipment to an}' lot which contained " not more than 15 per cent of wormy or weevil-infested beans" (5). The 15 per cent limit is tentative only. Acting under this ruling, a shipuient of broad beans from a San Francisco broker was seized in New York, after having been found to be over 25 per cent infested. The broker wa:s tried for violating the Food and Drugs Act, convicted 1 Numbers in parentheses refer to " Literature cited," p. 22. 132902°— 19— Bull. 807 1 2 BULLETIN 807, U. S. DEPARTMENT OF AGEICULTURE. by a jury, and fined $150 (9). Since then many dealers have refused to handle horse beans, while those who continued to deal in them have had to exercise caution that no shipment contained more than 15 per cent infested beans, or else run the risk of confiscation. The numerous confiscations during the past few years of shipments in transit by food inspectors, and the cost of hand picking to keep the infestation within the 15 per cent permitted, has resulted in keep- ing the price of beans low and reducing the acreage. Although the bean is commonly called the horse bean, the name is somewhat a misnomer, as only about 30 or 40 per cent of the crop is used as stock feed. The larger portion is shipped to New York and other eastern cities, where it is used for food by Italians and Portuguese, and is known as fava. Other names are English bean, Windsor bean, and tick bean. The horse bean is used also as a green vegetable, and of late years has been planted to a considerable extent as a winter cover crop, especially in fruit orchards. A recent Farmers* Bulletin (8) recommends further plantings along the Pacific coast, the Gulf of Mexico, and the South Atlantic States, not only for the dry beans for '^ \ j human consumption, but also for the green vege- table, stock feed, and green manuring. DESCRIPTION. ^ "/W//f/" THE EGG. Fig. 1.— The broad- The egg (fig. 1) is elliptical-ovate, about twice as bean weevil : Egg. • 1 + -fl • ^ ^ ^ • 1 -^ Oxreatiy en- long as "Wide, a trifle more pointed anteriorly, its larged. (Chitten- surface smooth with no visible sculpture. ^^Hien first laid it is whitish and glistening, but gradually turns darker. Just before hatching, the black head of the embryo can be seen plainly through the shell. THE LARVA. The young larva is pale yellow, with dark or black head and mouth- parts. The full-grown larva is cream colored, with small brown head and black mouth-parts. It is 4.5 to 5.5 mm. long and 2.5 to 3 mm. wide. THE PUPA. The pupa when first formed is light yellow or cream colored, with legs and wing-covers whiter. It gradually turns darker, particularly the appendages, until it is light brown before transforming. It measures about 3 mm. wide and 5 mm. long. THE BROAD-BEAN WEEVIL. THE ADULT. The adult (fig. 2) is from 3.5 to -1.5 mm. long and a little over half as wide. The general color is black, with white markings on the elytra and pygidium, giving it a somewhat mottled grayish appear- ance. The head is dark. The basal four joints of the antennie are reddish brown, tlie remainder black. The forelegs are reddish brown and black, while the middle and hind pairs are black. The species closely resembles the pea weevil (Z?. pisonim L., fig. 3), but may be separated by the following characters : Posterior femora acutely dentate; thorax broad; pattern of elytra well defined ; pygidium with a pair of dis- tinct apical black spots pisorum L. Posterior femora obtusely or obsoletely dentate ; thorax narrow ; pattern of elytra more or less suffused ; pygidium with black apical spots lacking or illy defined mfimanus Boh. Fig. 2. — The broad- bean weevil {B ru c h u s mfi- manus) : Adult or beetle. Enlarged. (Chittenden.) SYNONYMY. Bnichus rufimanus Boheman. BrucJms rufimanus Schoenlierr, Qen. et Spec. Curculionidum, v. 1, p. 58, 1833. Bnichus granarius auct. (not L.) Westwood, Curtis, Ormerod, Wood, Riley, Fletcher, Lintner, et al. Mylabris rufitnaiia Boh., Baudi, Deutsch. Ent. Zeitschr., ISSO, p. 404. The Bruchus granarius L. is Laria atomaria L., Syst. Nat., 12th ed., p. 605, 1776-1778. RECORDS OF OCCUR- RENCE IN CALIFOR- NIA. Although horse beans were grown in Cali- ornia as early as 1887, the horse-bean weevil was not recorded as ac- tually established in the United States until September 18, 1909. On that date Mr. I. J. Condit, then collaborator of the Bureau of Ento- mology, collected live specimens at San Luis Obispo, on growing horse beans (2, 3). Inquiry by the writer among buyers and growlers indi- cates that the weevil was established in California many years before 1909. Mr. P. G. Hammer, San Francisco, writes as follows, "We are quite positive of the date, for in our dealing with this variety of beans the first indication of weevil infestation appeared possibly not later Fig. 3. — The pea weevil (Bruchus pisorum) : a. Beetle b, larva; c, pupa. Enlarged. (Chittenden.) 4 BULLETIN 807, U. S. DEPAETMENT OF AGEICULTURE. than 1888,'' from Alameda County. One other broker gives 1893 as the first date infested beans were observed by him from the same lo- cality. Mr. E. A. Blinker, from a personal transaction in 1898, posi- tively remembers that date as the first year infested horse beans were observed by him. Other dates for the different localities given by different bean brokers are as follows: Gilroy, 1890; Watsonville, 1900; Morro. 1900; Oceano, 1903; Halfmoon, 1904; and Sacramento, 1908. It must be remembered that in none of the above dates were the weevils in horse beans identified as Bruchus I'u-fiinanus^ but there is little doubt that this is what they were. It is, therefore, quite evi- dent that the insect was present in the different localities a number of years before September, 1909. DISTRIBUTION IN CALIFORNIA. The principal broad-bean sections in California are around San Francisco Bay and down along the coast to a little below San Luis Obispo, The insect is distributed all over this section, having been taken or reported from the following counties : Sonoma, Napa, Yolo, Sacramento, San Joaquin, San Mateo, Alameda, Santa Cruz, Santa Clara, San Benito, Monterey, and San Luis Obispo (fig. 4). Broad beans are grown in small quantities in many other counties, usually in back yard gardens or in small plots for the green beans. Of late years, in several counties, they have been quite extensively planted as cover crops, particularly in citrus orchards. Unless great care is exercised, which is not often done, in planting uninfested or treated seed, it is safe to say that the broad-bean weevil will be found wherever broad beans are grown and allowed to come to maturity. The writer has found eggs on the green pods of broad beans planted for cover crops in Los Angeles, Orange, and Eiverside counties. Cover crops, how- ever, are all plowed under before the weevil has had a chance to develop. SPREAD IN CALIFORNIA. The broad-bean weevil probably became established in Alameda County, where horse beans were first grown, about 1888. It next appeared in Santa Clara County in 1890, and in Santa Cruz, Santa Clara, and San Luis Obispo Counties in 1900. By 1904 it w\as reported from San Mateo County; by 1911 from Sacramento and Sonoma, by 1914 from Yolo and San Benito, by 1916 from San Joaquin, and by 1917 from Napa. DISSEMINATION. Although the adult insect is an active creature and doubtless can fly some little distance, from one field to another, or from where beans may be stored to a near-by field, the principal method of dis- THE BEOAD-BEAX WEEVIL. 5 semination is through the transportation from one locality to another of beans infested with live weevils. It is possible for the adnlts to escape both after the beans have reached their destination and while they are en route. Planting of these infested beans in a new locality is l)onnd to result in an infested crop. foci MOiUsismrou \ / "^ \MODtK ^0 -L tUMBT -I ^^ ^X-t J SHASTA lassch V \ ff"'"' /^ \ Xtchama A Ky ■ SIURA T OCIMO ) f\^ \5if#* ^ I C*ffl u /— V \ X^usa Kl " HAC£Rr Fig. 4. — Map of California, shov.ing counties from wliich tlie broad-bean weevil has been talccn or reported. NATURE OF DAMAGE. Damage caused by the broad-bean weevil is principally due to the fact that the larva feeds and transforms to the adult within the beans (PL I, fig. 3, «, h). Many adults remain in the beans for sev- eral months, and their presence renders the beans unfit for food. Some adults emerge soon after forming, leaving round holes in the 6 BULLETIN 807, U. S. DEPARTMENT OF AGRICULTURE. beans where the tissue was consumed by the hirvaB (PI. I, figs. 4, a, h, and 5) . This " buggy " appearance lessens the salability of the beans. Infested beans are not only reduced somewhat in weight (see Table I) but their germinating power is also lessened, as will be shown later on. Table I. — Shoicing amount of beans consumed by the broad-bean weevil. Number of -weevils in beans. Weight of 50 beans. Weight consumed. Percentage consumed. None Grams. 76,970 74, SOO 72,430 70,150 Grams. 1 2,170 4,540 6,820 2 81 2 6 89 3 8.86 This table .shows that approximately 3 per cent of the weight of a bean is consumed by the larva of each weevil developing in it. EXTENT OF DAMAGE. It is estimated that the area devoted to broad beans commercially is at present about 3,000 acres for the dry beans, 100 acres for can- ning green, and 1,000 acres for cover crops. In addition many hun- dreds of acres are grown in small lots, back-yard gardens, etc., for home consumption. The commercial production of dry beans for market is at present about 50,000 bags, but has been, before weevil damage became so general, as much as 200,000 bags. After a careful survey of the 1916, 1917, and 1918 crops of horse beans, and of numerous samples taken by food inspectors of the Bureau of Chemistry and the writer. Tables II and III were pre- pared. Table II. — Sumviarp of the percentage of infestation of the 1916, 1917, and 1918 crops of broad boans grown in California. Infestation. 0 to 5 per cent 5.1 to 10 percent. 10.1 to 15 per cent 15.1 to 30 per cent Above 30 per cent Total 1916 crop. 1917 crop. Number of bags. 18,851 8, 366 6, .835 11,114 4,318 49, 474 Per cent of total crop. 3S.10 16. 88 13.81 22.46 8.75 Number of bags. 15, 372 6, 855 4,904 4, 852 1,107 Per cent of total crop. 46.46 20.71 14.82 14.66 3.35 100.00 33,090 I 100.00 1918 crop. Number of bags. 6,638 5,341 4,629 10,547 3,681 33, 036 Per cent of total crop. 26. 75 16.16 14.01 31.93 11.15 100.00 THE BROAD-BEAN WEEVIL. 7 The damage according to the different horse-bean regions is shown in the following table : Table III. — Summary of the 1916, 1917, and 1918 crops of broad beans shoiving the percentage of weevil infestation by localities. 1916 infesta- tion. ft 1917 infesta- tion. 1918 infesta- tion. Sacramento: Per cent. 41 0 . 9.09 50 0 14.5 55 0 13.6 70 0 24 Per cent. 63 0 12.7 63 0 14.5 56.6 .5 10.9 46 1 16 Per cent. 84.3 1 22.4 Oceano-Morro: Maximum 17.2 0 Average 2.92 Halfmoon: Maximum ^. 56 0 Average . . . . . 16.8 Gilroy: Table II*shows that of the entire crop of broad beans for the three years 1916, 1917, and 1918, 31.21, 18.01, and 43.08 per cent, respec- tively, were above the 15 per cent limit of weevil infestation allowed by the Bureau of Chemistry, and therefore could not be shipped un- less hand picked. Table III shows that even the average percentage of infestation for the entire 1916 cix)p in the Halfmoon and Gilroy regions was above the 15 per cent limit, while the same is true of the Sacramento and Halfmoon districts for the 1918 crop. This table also shows some interesting data on the increase in infestation. Sacramento is a comparatively new district for raising broad beans, and at first the weevil infestation was low, but as planting continued from year to year, and the acreage increased, the percentage of in- festation greatly increased, going from a maximum of H per cent in 1916 to 63 per cent in 191T and 84.3 per cent in 1918, while the a.verage percentage of infestation for the same years increased from 9.19 to 12.7, and then to 22,4 per cent. Climatic conditions at Sacra- mento are adverse to late planting, and early planting as practiced there is favorable to heavy weevil infestation. Unless control meas- ures are practiced, and only uninfested or treated seed is planted, this district soon will be in the same position in regard to the production of horse beans that Alameda County has been in for some years. The low percentage of infestation in the Oceano-Morro district for the year 1918 will be explained later. Estimates from bean brokers on the reduction in value of broad beans from the infestation of the weevil vary from 25 per cent to total unsalability, depending on the degree of infestation. 8 BULLETIN 807, U. S. DEPARTMENT OF AGRICULTURE. The cost of hand picking to remove the Aveevily beans also de- pends on the degree of infestation, but it is estimated at about $1.50 per 100 pounds. Estimates of the reduction on the total horse bean acreage, because of the increase in weevil infestation, range from 25 to 75 per cent. Alameda County was formerly the largest producer, but at present, due largely to the extensive infestation of all horse beans grown there, practically no Joeans are produced commercially. It is also estimated that if weevil infestation could be prevented, the acreage for the dry- bean crop alone would be increased from 100 to 300 per cent in that county. It will be seen from Table III that there is considerable varia- tion in the percentage of infestation in the different districts. There is also great variation in the number of weevils developing in a single bean. In the case of the pea weevil B. fisorum (fig. 3), with a very similar appearance and life history, only one weevil develops in a seed, but with the broad-bean weevil there are often two and three adults in a single bean, while it is not at all rare to find four, five, and BA^en six. The following table illustrates this : Table IV. -Summary of the percentafie of broad beans infested with different number of weevils. Number beans examined. Locality raised. Year raised. Num- ber weevils per bean. 1 wee- vil per bean. 2 wee- vils ])er bean. 3 wee- vils per bean. 4 and 5 weevils per bean. Total per cent infested. 1000 Hayward do 1915 1916 1917 1917 1917 371 331 7S1 880 573 367 321 1(54 105 227 186 216 46 14 140 63 95 9 1 29 13 34 0 0 31 62.9 1 000 66. 9 1 000 do 21.9 I'ooo Stockton Hayward 12 1,000 42.7 58.9 23.7 12 3.9 1.5 41.1 LIFE HISTORY. The eggs are laid on the outside of the green pods, being cemented to the latter by a glutinous secretion. They are laid singly and in- discriminately over the surface of the pods (see PL I, fig. 1) without apparent reference as to whether the position is favorable or un- favorable to the newly hatched larva getting into the young bean. The number of eggs on a pod also bears no relation to the number of beans it contains, as often the former exceeds by many times the number of larvae which might develop therein. One pod 4 inches long was observed with 55 eggs deposited on it. The following table shows the variation in number of eggs laid on a pod. Bui. 807, U. S. Dept. of Agriculture, Plate The Broad-Bean Weevil- 1, Broad-bean pod showing eggs laid upon its surface; 2, broad beans showing entrance holes or "stings" of larvEe; 3, a, b, broad beans cut open to expose larvfe of the broad-bean weevil and the damage they have done; 4, 5, infested broad beans showing emergence holes made by adult weevils in leaving the seed. In figures, below the emergence hole, is shown a "window." THE BROAD-BEAN WEEVIL. Table V. — Number of eggs of Bruclms ruflmamis deijosited mi pods at Hay- ward, Calif. Date. Number ports on plant. Total number eg.cs per plant. Number eggs per pod. Maximum. Minimum. Average. April 20 1910. 12 19 5 12 14 8 8 11 10 88 153 30 70 119 71 48 115 144 17 17 18 29 22 19 9 23 42 2 1 2 1 1 4 3 2 8 7 3 Do g April 25 6 Do 5 8 Do 8 5 Do 8 8 Do 6 Do 11 3 Do 14 No eggs were observed by the writer except on the pods, the latter varying in size from less than an inch to full growth of 5 inches. Most of the eggs are found on the larger pods. Although adults have been noticed a number of times crawling over the plants during the dnj, the act of oviposition was never observed. It is probable that it occurs late in the evening or at night. Efforts to induce oviposition in captivity were not successful, so the total number of eggs laid by individual females was not ascertained. The duration of the egg .stage was found to be from 9 to 18 days, with an average of 13 days. A few days before hatching the dark head of the embryonic larva can be seen plainly through the egg- shell. The larva draws its head back, leaving about one-quarter of the shell hollow. For about a day the position appears to be un- changed, then gradually as the larva eats through the side of the shell attached to the pod and into the pod the clear space in the shell becomes larger until the latter is entirely empty, having a glass}^ and transparent appearance. It usually takes about tw^o days from the time the larva begins eating through the shell until it is entirely out of the latter and into the bean pod. The young larva begins feeding when the bean is green, and it is well along toward maturity before the bean dries up. The amount of food consumed is small compared with what other insects eat. It ap- l)ears that the young larva often eats out a short tunnel, advancing a little distance from the point where it entered the bean. When the larva gets larger it eats out a hole somewhat its own shape, and about 50 per cent larger, as is show^n in Plate I, figure 3, a and h. When the larva reaches maturity, and just before pupating, it eats out a round hole in the cotyledon, directly under the epidermis. The hole is plainly seen through the half-transparent skin or " window," which is broken easily by the adult weevil when it is ready to emerge (see upper beans in PI. I, fig. 5). The hole is seldom at the spot where the larva entered, but at varying distances from it. The 132902°— 19— Bull. 807 2 10 BULLETIN 807, U. S. DEPARTMENT OF AGRICULTURE. point where the young hirva entered, sometimes called the " sting," plainly shows on the dry bean as a dark spot varying in size from a pinpoint to almost as large as a pinhead. This is illustrated in Plate I, figure 2', a and h. Mortality among the larvae after entering the beans is rather high. This is so both in the case where hy far more larvae enter a bean than could develop in it, and in the case where only one or two larvae enter, although the percentage of mortality is less with the latter than with the former. Table VI graphically illustrates this point : Table VI. — Comparison of the number of larvw of Bruchus ruflmamis entering beans icitli the number reaching maturity. Number of beans. Number of larvae entering beans. Number of adults. Maxi- mum number of larvas entering a bean. Average number of larvse entering a bean. Average number reacliing maturity. Percent- age of larvfe entering beans reaching maturity. Percent- age of beans entered by larva. Percent- age of beans infested with adults. 100 263 199 226 72 54 132 95 86 16 29 10 7 16 9 5 2.63 1.99 2.26 .72 .54 1.32 .95 .86 .16 .29 50.1 47.7 38 22.2 53.6 85 78 80 40 39 70 100 63 100 58 100 15 100 26 Average 162.4- 71.6 9.4 1.62 .71 42.3 64.4 46.4 It will be seen from the table that although as many as 16 larvae may enter a single bean, the average number is about 2, and a total average of a little over 40 per cent reach maturity. Comparison with Table IV shows that while a maximum of from 5 to 16 larvae may enter a bean, the percentage of beans containing more than 2 adults is small. The larval stage is from 10 to 15 weeks, the average being about 12 weeks. Immediately after eating out the round hole in the cotyledon, already referred to, the full-grown larva becomes quiet, and in a short time pupates. At first the pupa is the same color as the larva, but gradually turns darker until it is a dark brown. The pupal stage is from 7 to 16 days with an average of 10 daj'S. When the adult is first formed it is light brown and very soft, but it gradually turns darker and becomes harder. The adult may soon eat its way out of the bean through the " window^ " prepared by the larva, or it may remain in the bean for several months. In fact many adults never emerge from the beans at all, but die in them. If the weather is warm, or the beans are handled much, the weevils are apt to emerge from the beans sooner and in greater numbers than if the weather is cool and the beans are not handled much. Many of the adults emerge from the beans after the latter have been planted. The duration of the adult stage varies considerably, de- THE BROAD-BEAN WEEVIL. 11 pending on the weather conditions, from one month to a possible eight months. Table VII gives a summary of the life-history records. Table VII. — Life history of the broad-bean iveevil in California. Locality. Eggs laid. Hatched. Dura^ tion of stage. Pupated. Dura- tion of larva stage. Adult. Dura- tion of pupa stage. Died. Durar tionof adult stage. Total length of life. Pasadena . Hayward . 1917. May 3 May 11 1917. May 15 May 24 Days. 12 13 1917. Aug. 19 Auj. 29 Sept. 3 1918. July 26 Aug. 5 Aug. 22 Aug. 5 ...do Aug. 22 Aug. 12 Sept. 10 Aug. 5 Aug. 30 Days. 96 98 '96' 80 94 97 106 64 87 1917. Aug. 27 Sept. 8 Sept. 12 1918. Aug. 4 Aug. 21 Aug. 30 Aug. 12 ...do Aug. 30 Aug. 28 Sept. 17 Aug. 13 Sept. 13 Days. 8 10 9 10 16 8 7 7 8 16 7 8 14 Sept. 27,1917 Feb. S, 1918 Dec. 7, 1917 Days. 31 152 86 Days. 147 273 1918. Apr. 9 do 1918. do Alhamhra. AUiamliia. Alham!ir;i. Apr. 28 May 6 May 7 Apr. 28 May 17 May 17 May 17 Mar. 14 May 5 May 17 May 20 May 7 May 27 June 2 June 4 Mar. 28 9 11 13 9 10 16 18 14 12.5 Apr. 23,19191 Dec. 26,1918 May 31,19191 253 137 274 359 235 393 Alhambra. Alhambra. Dec. 6, 1918 Nov. 30, 1918 114 78 203 197 90 10 140 258 1 Theoretical, showing possible life history from two records. SEASONAL HISTORY. Egg laying begins about the middle of March, is heaviest in April, and extends a little beyond the middle of May. The larval period is from the latter part of March to the middle of October. Pupse can be found from the first of August to the latter part of October, and adults from the middle of August to about the following June. Tables VIII and IX show the actual records of the seasonal history for 1917 and 1918. Table VIII. -Sca.^onal histtonj of Bruchus rtifimanus in Pasadena and Alhambra, Calif. Year. First eggs ob- served. Last eggs ob- served. First eggs hatched. Last eggs hatched. First full- grown larvEe. Last full- grown larvffi. First pupa. Last pupa. First adult. La'?t adult. 1917 Mar. 23 Mar. 14 May 20 May 19 Aug. 1 July 16 Oit. 19 Sept. 26 Aug. 3 July 18 Oft. 22 Sept. 28 Aug. 10 July 25 (1) lyl.S Mar. 28 June 4 1 About June 1, 1918. 12 BULLETIN 807, U. S. DEPARTMENT OF AGRICULTURE. Table IX. — Grapliw chart of life history of Bruchus rufimanus for seasom of 1911-18 and 1918-19. Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. Nov. Dec. Eggs, Larvce, PupsR, Adults, - i 1 Continuous line=1917-18. Broken line=1918-19. Although a few adults may be found as late as June, the majority die before the end of March. During 1917 a number of badly in- fested beans were inclosed in a box. Ten per cent of the weevils were dead by February 1, 50 per cent by March, and 90 per cent by the 1st of April. A very few remained alive until June. The re- sults of field observations at Hayward during 1918 as to the preval- ence of weevils in the fields are recorded in figure 5. Each count represents an examination of horse-bean plants for a period of 15 minutes. /<3 ^^ / ilue-grass in waste places and along fences since pastures are kept cropped so closely that there is little chance for it to winter over there. It hibernates as a larva, pupates in spring, and the adult emerges early in May. It is one of the first species to make its appearance in the spring. Males normally occur. The ^gg is shown in figure 8 at e. THE FESTUCA JOINTWORM.i The Festuca jointworm was described recently by the writer and W. T. Emery (10, p. 454) from specimens reared from material col- lected by the writer near Youngstown, Ohio, in 1913. Since that date it has been located at other points in Ohio, New Jersey, Pennsyl- vania, and Virginia. It is the slenderest species among the gall formers, the abdomen being particularly long and narrow. MANNER OF INJUBY. E. festucae forms galls or hardened enlargements, usually above the second joint from the ground, although they may occur at any joint. The galls (PL VI, C) may be prominent or inconspicuous; in the latter case they can be detected only by pinching the stems between the fingers. The injury undoubtedly would be serious to the seed crop and very probably would shorten the hay crop, also, ns the flowering stalks are rather slender, and although not quite as frail, they are very much like blue-grass stems in that they are not woody. Festuca sp., therefore, would naturally suffer more from attacks of this kind than would orchard grass or timothy. HOST PLANTS. The writer has been rearing festucae from Festuca sp. in confine- ment and making observations in the field for 5 years and has not reared this species from any other host. 1 Harmolita festucae Phillips and Emery. JOINTWORM FLIES. 19 LIFE HISTORY. As this grass grows naturally in waste places and along fences its joint worm enemy does not depend upon cultivated areas to maintain itself. It winters in the old seed stalks as pupa, the adult emerging near the middle of May. The species is normally thelyotokous, males very rarely occurring. SPECIES INFESTING WILD GRASSES. In the j^receding pages the writer has given briefly some of the more important facts relating to all the species at present known to infest our grains and cultivated grasses. It may appear at first glance that wheat has rather more than its share of species, especially since it is our most important small grain. On the other hand, the genus Elymus, a wild grass, is the home of nine species, eight of which are gall-formers ; Agropyron sp., another wild grass, has four species, three of which are gall-formers. The species infesting the wild grasses will be dealt with very briefly. A good many of them may lay a just claim to some consideration from an economic standpoint, frofn the fact, as previously stated, that thej^ have parasites in com- mon with the more important economic species. HARMOLITA MACULATA Howard. Howard (8, p. 15) described this species as Isosoma maculatuTn in 1896 but nothing was known then of its life history. In re- cent years it has been collected by various members of the branch of Cereal and Forage Insect Investigations. The writer has reared it in confinement since 1912. It has persistently refused to breed in any host other than grasses of the genus Bromus. It does not form galls but lives in the walls of the plant stem, particularly of cheat {Broraus secalinus). The cheat stem is almost solid near the base, at least just above the lower joints, the walls being very thick. The Qgg (fig. 8, d) apparently is deposited in the walls of the stem somewhat like the ^gg of tritici and the larva excavates a little tunnel about half an inch long just above the joint (PI. V, B). There are sometimes two or three to an internode. It is a very widespread species and un- doubtedly occurs wherever cheat or other species of Bromus are found. There is only one generation a year.' The species hibernates in the larval stage, pupates in the spring, and the adults emerge in May. It is arrhenotokous under control conditions though in nature both sexes occur. HARMOLITA ATLANTICA Phillips and Emery. HarmoVita atJantica (10, p. 461) is a species which the writer first reared near Richmond, Ind., in 1909. A few specimens have been . 20 BULLETIN 808, U. S. DEPARTMENT OF AGRICULTURE. taken in Michigan and many in New York. It has been reared spar- ingly in confinement, not being a very tractable species. It has never been reared from any host other than Agropyron. This species forms galls in the stems. Some of the galls resemble somewhat those of tritici, being placed in the walls of the stem ; others resemble those made by vaginicola, occurring in the sheath surround- ing the head. In the latter case the head does not appear at all. (PI. VI, B.) There is only one generation a year, and, as with the majority of species, they pass the winter as larvse in the galls of the old plant stems, pupate in the spring, and emerge as adults in May. The spe- cies is normally thelyotokous, males rarely occurring. HARMOLITA AGROPYROPHILA Phillips and Emery. Harmolita agwpyropMla (10, p. 450) is the only species, besides grandis^ that has two generations a year. Singularly enough, Agro- pyron is one of the few plants infested by Harmolita that produce stems continuously throughout the growing season. The majority of the other jointworm species would find it difficult to maintain two generations a year on a single host unless they were dimorphic like grandis and one generation developed in the very young plants. H. agwpyrophila was collected first in 1904 by F. M. Webster. Since that time it has been collected by various members of the branch of Cereal and Forage Insect Investigations. The writer has been rearing and observing this species since 1905, when he first dis- covered that it has two generations in a year. The larvae inhabit the center of the stem and may be found at any joint. The species has never been reared from any host other than Agropyron although repeated attempts have been made to induce it to live on other plants in confinement. There are two generations a year, the first generation emerging very early, during the last week in April and the first week in May. It is one of the first species to emerge. The second generation emerges the latter part of June and the first week in July.^ The second generation deposits its eggs (fig. 2, (7) in the young tender stems, and as it is rather late in the season these stems do not produce heads. HARMOLITA ELYMI French. This species was described as IsosoTna elymi in 1882 by G. H. French (6) , who reared it from Elymus amerlcaniis from Carbondale, 111. The writer's attention was first attracted to it in 1908, when it was learned that it occupied the center of the stem of Elymus sp. ^ These dates apply to Richmond, Ind. Bui. 808, U. S. Dept. of Agriculture. Plate V. JoiNTWORM Flies of the Genus Harmolita. A, Characteristic gall of the rye jointworm (//. secalis) in rye (about natural size); B, larvae of H. maculata in a cheat stem (about natural size); C, galls of H. elymivora m Elyraus sp. (about natural size j. Bui. 808, U. S. Dept. of Agriculture. Plate VI. JoiNTWORM Flies of the Genus Harmolita. A, Characteristic galls of //. elymicola in Elymus sp., slightly reduced; B, two types of injury to Agropyron sp. by H. allantica, those to the left being sheath galls and the long stems to the right being stem galls, slightly reduced; C, Typical galls of the Festuca jointworm (H. festucae), about natural size; D, E, characK-ristic galls of H. elyniophthora, en\siTged shghi\y . (Photographs D and E by Mr. C. N. Aiuslie.) JOINTWORM FLIES. 21 Nothing was known of its life history until recent years. It has a wide range, being found as far west as Utah, and having been reared from collections of Elymus from Illinois, Indiana, Ohio, and Vir- ginia. It probably occurs wherever Elymus sp. grows normally. H. elymji inhabits the center of the stem of Elymrms sp. and has never been reared from any other plant, although repeated attempt's have been made to rear it from wheat, rye, barley, and some of the grasses. There is only one generation a year. Hibernation is in the larval stage, pupation occurs in the spring, and the adults emerge in May. This species is thelyotokous, males never having been observed. HARMOLITA ELYMICOLA Phillips and Emery. Harmolita elymicola (10, p. 460) is the commonest species on Ely- mus in the Eastern States and is apparently a strictly eastern species, as it has not been collected west of the Mississippi Eiver with the ex- ception of southeastern Missouri. East of the Mississippi it has been taken in Illinois, Indiana, Michigan, Ohio, and Virginia. It forms very prominent galls or enlargements, usually above the second joint from the base of the plant (PI. VT, A) . It has never been reared from any host other than Elymus. Repeated attempts have been made to breed the species on wheat, barley, and various grasses. It has been observed to oviposit in wheat and barley stems, but noth- ing ever developed in these stems. Figure 8 c, shows eggs of this species. II. elymicola has only one generation a year. It hibernates as a larva in the old seed stalks, pupates in the spring, and emerges in May. It has proven to be arrhenotokous under control conditions, but both sexes normally occur in nature. HARMOLITA ELYMIVORA Phillips and Emery. Harmolita elymAvora (10, p. 464) is not nearly as abundant as elym/icola, though it apparently has a wider range, having been found as far west as Arkansas. It has also been reared from stems of ElymMs sp. collected in Indiana, Michigan, Ohio, and Virginia. H. elymivora forms galls in the stems of Elymus sp. just below the head. As a result, the head or fruiting body of the plant never develops (PL V, C). It is possible that it forms galls of the elymi- cola type also, but the writer has never reared any from such galls. The life history is similar to that of elymicola. Eggs are shown in ligure 2, c. HARMOLITA RUFIPES Phillips and Emery. IlarTnolita ruftpes has been described only recently (10, p. 453), but F. M. Webster is probably the first to have reared it. In Bulletin 42 of the Division of Entomology Prof. Webster confused this species 22 BULLETIN 808, U. S. DEPARTMENT OF AGRICULTURE. with H. hordei Harris, stating that he had reared hordei from Elymus at Champaign, 111. The writer has examined these speci- mens and they are very clearly rujifes. The species resembles hordei somewhat in that the legs are reddish brown, but it is much larger than hordei and other characters very readily distinguish them. Rufifes is apparently a strictly western species, never having been reared any farther east than Champaign, 111. It has been reared from Elymus sp. collected in Kansas, Nebraska, New Mexico, and Utah. It has never been reared from any grass other than Elymus sp. Repeated attempts were made by the writer to rear it in confine- ment at both La Fayette, Ind., and Charlottesville, Va.,upon its host, but it has steadily and persistently refused to breed. It apparently would oviposit, but no larvae ever developed. It forms galls in the stems of Elymus sp. Nothing is known of its life history further than that it has a single generation during the year, hibernating in the usual way and emerging in May. Both sexes occur. HARMOLITA HESPERUS Phillips and Emery. HarTnolita hesperus (10, p. 457) was considered by the writer for quite a while to be rufi-pes. The two species can be distinguished, however, very readily. H. hesperus is apparently a strictly western species, having been reared from Elymus sp. collected from Kansas and Utah, but it has never been found east of the Mississippi Eiver. H. hes'perus forms galls very much as does elymlcola. It has not been reared from any plants other than Hordeum juhatum and Elymus sp. Nothing is known of its life history further than that it hibernates in the usual way and emerges in May. Both sexes occur normally. The writer has never been able to rear this species in con- finement, though repeated attempts have been made to rear it in Elymus sp., both at La Fayette, Ind., and at Charlottesville, Ya. The e,gg is shown in figure 7 at a. HARMOLITA ELYMOPHTHORA Phillips and Emery. TJannolita e^ymofhthora (10, p. -105) appears to be a strictly west- ern species also, having been reared only from Nebraska and North Dakota. It was first brought to the writer's attention by C. N. Ains- lie, who sent in galls on Elymus sp. from which was reared this species. It forms galls in Elymus sp. (PI. VI, D, E), though it is not Imown whether it has other hosts or not. It has refused to breed in confinement at Charlottesville, Va. It has one generation a year and both males and females normally occur. HARMOLITA OVATA Phillips and Emery. Earmolita ovata (10, p. 458) has been reared only from Kansas. It was sent to the writer in the fall of 1914 by E. O. G. Kelly, who was JOINTWORM FLIES. 23 then a member of the Cereal and Forage Insect Investigations force. It forms galls in Elymus sp. It is not known whether it has other hosts. It refused to breed in confinement at Charlottesville, Va. There is one generation a year and apparently both males and females occur. SPECIES WHOSE BIOLOGY IS UNKNOWN. The species previously treated in this paper have all been reared repeatedly in cages under artificial conditions with the exception of captiva^ rufipes^ Jiesperus^ elymopKtTwra^ and (yvata. The writer has never seen living specimens of the remaining six species described by Phillips and Emery (10) — poophila^ agropyrocola^ occidentalism elymophila, elyTnoxena^ and gillettei — or of hromicola Howard and agrostidis Howard, and practically all that is known concerning them is contained in the meager data incidental to collection. Four of these, agrostidis^ hromicola^ elymoxena^ and elymophila^ are from California; two, gillettei and poopMla^ are from Colorado; one, agropyrocola^ from Utah ; and occidentalis from New Mexico. Noth; ing, of course, is known of their life histories. Poophila was reared from Poa lucida^ sent in by A. D. Hopkins from Husted, Colo.; agropyrocola and occidentalis were reared from Agropyron sp., the former reared from material sent in by Desla Bennion and the latter from material forwarded by Y. L. Wildermuth ; hromAcola was reared from Bromus ciVmtus; agrostidis was reared from Agrostis sp. ; elymopMla and elymoxena were reared from Elymus sp. The four last species were collected by Albert Koebele. H. gillettei, as the name implies, was reared by C. P. Gillette and was named for him; the host is unknown. Undoubtedly further observations and collections will add many more new species from the Western States, and more particularly from the Pacific Coast States. CONTROL MEASURES. Farmers, as well as entomologists, have concerned themselves very little about controlling these really serious pests. Fortunately or un- fortunately, depending upon the point of view, the parasites have taken care of the situation to such an extent that only now and then the jointworm {11. tritici) gets out of hand and causes the almost total destruction of a crop in a given locality. Therefore the matter has been viewed very calmly and a toll of from 1 to 5 bushels or more per acre has commonly been tolerated. We have been perfectly content to pay an annual tribute in preference to fighting vigorously to throw off this burden. But for the parasites we should have been obliged to bestir ourselves long ago or else abandon wheat growing in the Eastern States. There is some excuse for this condition of 24 BULLETIN 808, U. S. DEPARTMENT OF AGRICULTURE. affairs, in that there was no very simple remedy at hand ; for example, simply spraying the wheat would not cure the trouble. The writer has studied the situation for years and has found that most if not all of the so-called remedies are useless. For several winters futile attempts were made to bum over stubble fields in Indiana. In the first place the ground must be frozen or the burning will destroy the young clover and at no time during the winter could the stubble be burned under these conditions. This would be the simplest of all remedies if it were practicable, although it would destroy a source of humus. Another supposed remedy was to plant wheat as far from the old stubble as possible. This is folly unless the farm in question covers an area of several square miles. The jointworm can fly and may be carried by the wind at least for a mile. The suggestion of sowing wheat early, in years of scarcity of the Hessian fl}^, is equally futile. V. L. Wildermuth, of this bureau, made observations on 51 fields at Groveport, Ohio, in 1909, in order to determine the effect of early and late sowings in relation to joint- worm injury. Twenty-eight fields sown from September 10 to 30, with an average date of September 24, showed an average infestation of 38 per cent, with an average yield of 8| bushels per acre; 15 of these fields produced grain of good, 10 of medium, and 3 of poor quality. On the other hand, 23 fields sown October 1 to 24, with average date of October 10, showed an average infestation of 19 per cent with an average yield of 20 bushels per acre; 11 of these pro- duced grain of good, 9 of medium, and 3 of poor quality. These observations show also that whether the wheat were sown early or late the wheat on poor ground always was infested to a greater degree, provided the fields were equally distant from the source of infestation. The writer's observations confirm this idea. Some writers have suggested that the larvae would be destroyed if the infested stubble should pass through stables and then be sub- jected to the process of decomposition incidental to composted ma- nure. The writer therefore placed infested stubbles in horse stables and allowed them to become mixed with the droppings in the ordi- nary way. They w^ere then removed to the center of a heap of manure and part of them allowed to remain in the manure all w^inter while the remainder were removed after having remained in it for two weeks, and then scattered on the ground and allowed to remain there undisturbed throughout the winter and spring. Those remain- ing in the manure pile were removed the last week in March, and placed in confinement for observation, as were those that were scat- tered on the ground in early winter. Each lot produced about the same number of adults as did stubble that was allowed to winter in the usual manner in the field. This experiment certainly indi- JOINTWORM FLIES. 25 cates that there is small hope of destroying the joint worm by passing infested straw through the manure pile. The writer has been experimenting for the past three years in plowing under wheat stubble for the x^urpose of observing the effect on infestation the following year in wheat planted on the same ground. This work was done in the vicinity of Charlottesville, Va., where no infestation occurred nearer than 5 or 6 miles. There- fore the data secured should be reliable, as there was no opportunity for the jointworm to come in from neighboring fields. Small plats were used and the infestation was accomplished by bringing in a very large amount of infested stubble from a distance, placing it in the growing wheat, and allowing the adults to emerge and infest the plat. In order to insure heavy infestation, large amounts of badly infested stubble were brought in again from a distance after the wheat was cut, making accurate counts of the number thus brought in, and then carefully estimating the infestation that already existed on the plat. The total number of wheat stems or stubbles was then counted on several representative square yards of the plat. In this way the percentage of artifical infestation could be deter- mined accurately. The infested stubbles which had been introduced, together with the stubbles already standing on the plat, were then turned under. The ground was plowed as soon after harvest as possible, disked, and some crop like peas or soy beans sown. The plat was disked again in the fall at the proper time and reseeded to wheat. In this way an 8 per cent infestation was reduced to 1 per cent the following year. A second trial reduced the infestation from 32 per cent to 3 per cent. A third trial in 1918 reduced a 19 per cent infestation to 2 per cent. These experiments indicate that plowing under stubble is a very effective remed}'. It doubtless would destroy all of the insects if all of the infested stubbles could be completely buried, but it is im- possible to do this. Nevertheless it is practicable to control the ravages of the species in this manner. "W^iile this method of con- trol would necessitate a change in the existing rotation of crops where wheat is used as a nurse crop for clover, it would seem that it should be adopted if millions of dollars could be saved every year in this manner. Some agronomists admit that it would be prac- ticable to change the existing system of rotation so as to permit plowing down stubble to suppress important insect pests. If this were done not only the jointworm but also the Hessian fly would be controlled, and thus two of the major insect pests of wheat would be largely shorn of their power for harm to our most valuable bread grain. 26 BULLETIN 808, U. S. DEPARTMENT OF AGRICULTURE. The barley jointworm, the rye jointworm, and the rye straw- worm undoubtedly can be controlled in the same manner, if the necessity arises. The wheat straw-worm is very easy to control. Since one genera- tion is wingless it is only necessary to keep down all volunteer Avheat and never plant wheat nearer than 40 to 50 yards to infested stubble. This is allowing a wide margin of safety since they are supposed to be able to travel at most only 12 or 15 feet. Should it become necessary to control the straw- worms and joint- AYorms infesting our cultivated grasses, such as timothy, orchard grass, blue-grass, and Festuca sp., this undoubtedly could be accom- plished by clipping such fields in the spring in order to delay the appearence of seed stalks until the emergence of the insects, when there would be no place for them to deposit their eggs. JOINTWORM FLIES. 27 LITERATURE CITED. (1) ASHMEAD, W. H. 1S94. Descriptions of new parasitic Hymenoptera. In Trans. Amer. Ent. Sec, V. 21, p. 329. (2) DOANE, R. W, 1916. A new species of Isosoma attacking wheat in Utah. In Jour. Econ. Ent., V. 9, no. 4, p. 398-401. (3) Fitch, A. 1859. A new barley insect. In Jour. N. Y. State Agr. Soc, v. 10, p. 114- 115. (4) 18G1. The new insect in rye. In Amer. Agr., v. 20, p. 235-236. (5) 1862. Sixth report on the noxious and other insects of the State of New York, p. 154. Albany. (6) French, G. II. 1SS2. Two new species of Isosoma. In Can. Ent., v. 14, p. 9-11. (7) Harris, T. W. 1830. Insects. In New England Farmer, v. 9, no. 1, p. 1-2. (8) Howard, L. O. 1896. The grass and grain joiutworm flies and their allies. U. S. Dept. Agr. Div. Ent. Tech. Ser. 2. 24 p. (9) LiNTNER, J. A. 1S8S. Fourth report on the injurious and other insects of the State of New York, p. 34. Albany. (10) Phillips, W. J., and Emery, W. T. 1919. A revision of the North American species of the genus Harmolita north of Mexico. In Proc. U. S. Nat. Mus., v. 55, no. 2281, p. 433-471, pi. 39-48. (11) Riley, C. V. 1882. A new depredator infesting wheat stalks. In Amer. Nat., v. 16, p. 247. (12) — 1884. A new insect injurious to wheat. In Bui. Brooklyn Ent. Soc, v. 7, p. 111-112. (13) Webster, F. M. 1892. Some features of apparent jointworm attack. In U. S. Dept. Agr. Div. Ent, Ins. Life, v. 5, no. 2, p. 89-90. (14) 1903. Some insects attacking the stems of growing wheat, rye, barley, and oats. U. S. Dept. Agr. Div. Ent. Bui. 42. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 5 CENTS PER COPY V UNITED STATES DEPARTMENT OP AGRICULTURE BULLETIN No. 809 Contribation from the Barean of Entomology L. O. HOWARD, Chief Wasliington, D. C. PROFESSIONAL PAPER March 10, 1920 AMERICAN FOULBROOD By G. F. WHITE Specialist in Insect Diseases CONTENTS .Page Introduction ......•• 1 Name of the Disease 2 Healthy Brood &t the Age at Which It Dies of American Foulbrood . . 3 Symptoms 6 Etiology ....11 Technique 17 Thermal Death Point of American Foulbrood Spores 22 Resistance of American Foulbrood Spores to Drying 29 Resistance of American Foulbrood Spores to Direct Sunlight ... 29 Resistance of American Foulbrood Spores to Fermentation .... 31 Resistance of American Foulbrood Spores to Chemical Disinfectants . 31 Effect of Drags on American Foulbrood .32 Modes of Transmission 34 Diagnosis . . 36 Prognosis 38 Summary and Conclusions 39 Literature Cited ...42 Explanation of Plates ............ 45 WASHINGTON GOVERNMENT PRINTING OFFICE 1920 UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 809 ^^ Contribution from the Bureau of Entomology ^^vFv^^ SJ^'^^^U ^- ®- HOWARD, Chief ^^''^U Washington, D. C. PROFESSIONAL PAPER March 10, 1920 AMERICAN FOULBROOD By G. F. White Sj^ecialist in Insect Diseases CONTENTS Introduction - • Name of the disease Healthy brood at the age at which it dies of American foulbrood Symptoms Etiology Technique Thermal death point of American foulbrood spores Resistance of American foulbrood spores to drying Resistance of American foulbrood spores to direct sunlight Page 1 2 Resistance of American foulbrood spores to ferment ation 31 Resistance of American foulbrood spores to chemical disinfectants 31 Effect of drugs on American foulbrood 32 Modes of transmission 34 Diagnosis 36 Prognosis 38 Summary and conclusions 39 Literature cited 42 Explanation of plates 45 INTRODUCTION American foulbrood is an infectious disease of the brood of bees caused by Bacillus larvae. The disease is characterized by a decided ropiness of the decaying brood and a peculiar foul odor. It is very widely distributed, is readily recognized, and is of much economic importance. Its existence has been known for a long time, and bee- keepers have established many facts concerning it through observa- tions made while practicing their profession. While there is a con- sensus of opinion among beekeepers on certain points regarding the disorder, there are others on which a diversity of views has existed and still others about which almost nothing of a definite nature has been known. Former work was directed primarily toward the determination of the cause of the disease. Among the problems considered in the present studies are : The resistance of Bacillus larvae to heat, drying, sunlight, fermentation, and disinfectants; the effect of the disease on the colony and on the apiary; and the transmission, diagnosis, and prognosis of the disease. Direct studies on the treatment of the 132862°— 19— Bull. 809 1 2 BULLETIN 809, U. S. DEPARTMEISTT OF AGRICULTUKE. disease have not been attempted by the writer. It will be readily recognized, however, that any treatment that is efficient and at the same time economical must be determined by results obtained by the solution of such problems as those which have received attention in these studies. The facts may tend to engender fear for the disorder in some in- stances while in others they may tend to allay it and to offer encour- agement. It is hoped, however, that no statement made here will cause any beekeeper to lessen the vigilance that the disease requires, nor, on the other hand, to increase it to a point that would render its control uneconomical. The discussions in the present paper are based almost entirely upon observations made in the laboratory and in the experimental apiary. The value of the results is emphasized by the fact that the disease produced experimentally and the disease encountered in nature are identical in almost every respect. It is believed that the paper^ will be of interest not only to the practical beekeeper who wishes to apply the results noted here in the practice of his profession, but also to those who may desire to make further studies on the disease. NAME OF THE DISEASE That bees suffer from diseases is recorded in works written before the Christian era but it is not altogether clear what the diseases were. In 1771, Schirach (19)- was using the term ''foul brood" for an abnormal condition of the brood of the bees, but from his conception of the cause of the disorder one is led to believe that more than one abnormality was being referred to by the term. In 1882 Dzierzon (11) had definitely concluded that there were two kinds of foulbrood. Cheshire expressed a similar belief in August, 1884, but by September he had reached the conclusion that there was but one. In 1885 Cheshire and Cheyne (9) published an article containing the results of some studies on foulbrood including a description of Bacillus alvei. For more than a decade after the appearance of the paper, the view was cjuite generally accepted that there was but one disease present in the condition that was being called foulbrood and that B. alvei was the cause of it. Many American beekeepers, those in New York State especially, became convinced, some time during the decade from 1890 to 1900, that two serious brood diseases were being referred to hy the one name — foulbrood. That there are two such diseases has been conclusively proved. In the United States the one characterized by a decided ropiness of 1 The studies reported in the present paper are similar in nature to those made by the writer on sacbrood (25), Nosema-disease (26), and European foulbrood (27). These papers may be helpful where the dis- cussions in the present one are especially brief. The investigations were completed in September, 1916, and the paper was submitted for publication in October, 1918. s Figures in parenthesis refer to "Literature Cited," p. 42. AMERICAN FOULBROOD. 3 the decaying brood and a peculiar foul odor is now called American foulbrood, and the other one which is not so characterized is called European foulbrood. These two foulbroods are very different, the principal point of similarity being that they are both brood diseases. Both of them occur in Europe as well as in America. Unless these facts are borne in mind the names are likely to be misleading. The term "foulbrood" (French, la loque) in most countries as in America frequently is used in a general sense meaning simply some disorder of bees but no definite disease. In this popular use of the term, either or both of the two foulbroods may be meant. Other brood disorders sometimes are loosely referred to by this general term, "Foulbrood" and "bee pest" or slight modifications of these terms as used in different countries certainly include the disease American foulbrood. In Switzerland no pronounced odor had been observed in connection with American foulbrood and it is referred to as "nichtstinkende Faulbrut" (6, 7, 8); in Austria it is called "Faulbrut" (18); in Germany "Brutpest" (29) and "Faulbrut" (12) ; in Denmark it is called "Bipest" (2, 3, 4) ; in England (10) and in Ireland it is called foulbrood or bee pest; and in Australia it is called foulbrood (5). There are at least three infectious brood diseases of bees but as a rulj the one which beekeepers, entomologists, and pathologists have referred to in the past by the term "foulbrood" is the disease Amer- ican foulbrood, discussed in the present paper. HEALTHY BROOD AT THE AGE AT WHICH IT DIES OF AMERICAN FOULBROOD The description of the symptoms of a brood disease as weU as the recognition of the disease are very materially aided by making a comparison of diseased with healthy brood. Such a comparison involves the age of the brood, the relative arrangement of the capped and uncapped cells, the appearance of the diseased and of the dead brood, and the relation of the dead larva or pupa to the cell in which it lies. Furthermore, the character of the caps and the odor of the brood-combs should not be overlooked. In healthy brood-combs a certain regularity is to be expected in the arrangement of areas containing eggs, larvae, pupae, and emerg- ing brood, respectively. The cell caps when recently constructed are convex outward (PI. II, A), but usually become flattened some- what before the bees are ready to emerge. They are rarely punc- tured, remaining as a rule entire. In color these combs vary widely from a very light hue when recently constructed to a dark brown when old. Accompanying them is a slight but not disagreeable odor. In American foulbrood the brood that dies does so nearly always either during the last two days of the four-day prepupal period or 4 BULLETIN 800, U. S. DEPARTMENT OF AGRICULTURE. the first two days of the pupal period. Such brood, therefore, is in capped cells. A description of a healthy larva at this age is given by the wi-iter in a paper on sacbrood (25) and will be recounted here only briefly. A larva (prepupa) (PI. II, D, G; PI. VI, G), at the age at which death from American foulbrood takes place as a rule, lies extended lengthwise in the cell with its dorsal side against the floor. It is motionless with its head in the direction of the mouth of the ceU, its extreme anterior end extending nearly to the cap at the angle fonned by the cap and the roof. Its posterior third lies upon the bottom of the cell and extends to the roof; its length is approxi- mately that of the cell, being about one-half of an inch; its width is that of the cavity of the ceU, about one-fifth of an inch; and its two lateral sides cover about one-half each of the two lateral walls of the cells. The ventral surface of the larva is in general convex from side to side and concave from end to end. Transverse ridges and furrows give a segmented appearance to the body. An empty space of considerable extent occurs between the ventral surface and the roof of the cell (PI. VI, G). The color of the larva at this tunc is ahnost white with a slight bluish tint. The entire surface is glistening. The body wall is sufficiently resistant to permit the removal of the larva intact if care is exercised. The tissue mass as seen when the body wall is ruptured is semiliquid and nearly white in appearance. Upon microscopic examination it is found to be made up very largely of fat cells, and to be free from bacteria and other microorganisms. The change from the larva to the pupa (PI. VI, B), as far as out- ward appearances are concerned, takes place rapidly. Infrequently death from the disease takes place during this short period. At this time the bee (PI. VI, B) has neither the outward form of a larva nor of an adult. This stage of the traAsfomiation is recognized by the fact that the head is smaller than that of the complete pupa and the appendages are largely wanting. The pupa (PI. IV, A, D) immediately after transformation is sunilar in form and size to the adult bee. It rests with its back against the floor of the cell and with its antemise, proboscis, and legs lynig along its ventral surface. The posterior third is more pointed and much smaller than the same third of the larva. It lies upon the bottom of the cell but does not extend to the roof. The color of the pupa at the time of its transformation is nearly white with a slightly bluish tint. The first pigment observed is seen in the compound eyes. Pigmentation of other parts of the body follows. When death from American foulbrood takes place, it does so almost invariably before the time that pigmentation of the body other than of the eyes has occurred. The body wall of AMERICAN POtTLBROOD. 5 the healthy pupa for the first two days of this stage is easily ruptured. Its tissues are soft and when crushed are creamlike in consistency. During this period some care must be exercised in removing the body intact. Soon the body wall toughens so that the form is then maintained rather rigidly. SYMPTOMS Much of our knowledge concerning the symptoms of American foulbrood has been gained through observations made by beekeepers while practicing their profession. The apiary in which the disease has been produced by experimental inoculations offers an oppor- tunity to obtain a fairly complete picture of the disease. The present discussion of symptoms of the disorder is based upon observa- tions made on the disease thus produced. During these studies it has been possible to duplicate observations already made by beekeepers on the disease as it occurs in nature and to make stiD others. It is quite probable that a number of these additional obser- vations could be' duplicated on the disease in nature if a sufRciently close study were made. GENERAL SYMPTOMS In American foulbrood the symptoms vary within wide limits. The colony may or may not be noticeably weakened. If recently infected the strength will not be affected appreciably, but if the infection has been present for a considerable period the colony will be weakened as a rule. Not infrequently during the course of the disease the strength of the colony is diminished and death is the result. Between these limits any degree of strength may be encountered. The occasional larva which dies of American iFoulbrood (PI. VI, A) before it reaches the age at which healthy larvse are capped rarely, if ever, is capped after its death. Aside from these few larvae the brood that dies of American foulbrood does so in capped cells. After death the caps are removed by the adult bee from a large but varying proportion of cells containing dead brood (PI. Ill, C, F, I; PI. V, B), although they are allowed to remain on a consid- erable proportion of them (PI. VI, C, E, F; PI. II, B, C, E, F, H, I; PI. Ill, A, B, D, E, G, H; PL IV, B, C, E, F; PI. V, A, C, D, E, F). The irregular condition of the brood comb sometimes referred to by beekeepers as the "pepperbox" appearance (PI. I, A, B, C) is due very largely to the presence of cells containing dead brood which have been uncapped by the bees, occurring among similar cells which have not been uncapped, together with cells containing healthy brood either capped or uncapped. The caps over dead brood (PI. II, B) appear in many instances similar to those covering healthy brood. In many instances, 6 BULLETIN 809, U. S. DEPAKTMEKT OF AGKICULTURE. however, they are altered, being sometimes punctured (PI. Ill, A; PL VI, C), sometimes sunken, and sometimes darkened. Sometimes these abnormal appearances are combined. The holes in the caps vary in number and dimensions. Ordinarily there is but one (PL III, A), although there may be two (PL VI, C) or more. They are usually the size of a pinhead or smaller. The puncturing of the caps is done by the adult bees, and at the time of observation any proportion of the cap may have been removed. Sunken caps are found only after the disease has been present in the colony for a considerable period, weeks at least. This symptom is most marked after the brood frames have been roughly handled, as in shaking bees from them or shipping them by express or mail (PL I, C). This condition of the caps is encountered more fre- quently in samples shipped to the laboratory than in those taken directly from the affected colony. By rough handling the viscid decaying mass within the cell is brought in contact with the cap, adheres to it and tends to draw the cap inward as it settles, accounting in a large measure for the condition referred to as sunken cappings. The dark caps occur somewhat later and also depend largely upon the presence of decaying brood material within the cell. The caps of many cells containing dead brood, however, are neither simken nor darkened by its presence. When recently dead the decaying brood in American foulbrood is light brown in color, the shade deepening as the process of decay continues. The color passes through a chocolate, coffee, and mahog- any brown, reaching a very dark shade as it approaches the scale stage. The body wall of a larva or pupa dead of the disease soon softens and is easily ruptured, making it impossible to remove the remains intact from the cell. As the process of decay continues the mass becomes viscid and before the scale stage is reached the viscidity is such that it is capable of being roped out into fine threads to a dis- tance of 2 or 3 inches and sometimes even more. The scale when dry adheres more or less firmly to the cell wall. The odor of the brood combs, when it is present, is character- istic of the disease and may be spoken of conveniently as the foul- brood odor. It is never detected in the early stages of the disease and if only scattering cells are present seldom is noticed. When much brood is dying and the disease has been present for wrecks, the odor is noticeable and sooner or later becomes marked. When marked it is recognized by its strong and penetrating character and is usually thought of as being disagreeable. The brood combs after their removal from the hive tend to lose this foulbrood odor. The presence and extent of it in samples of the disease vary con- siderably, the variation depending largely upon the facts just mentioned. AMERICAN FOULBROOD. 7 It is possible to recognize most cases of American foulbrood from the general symptoms that are present. On the other hand many cases can be diagnosed only by an exact knowledge of the postmortem conditions of brood dead of the disease. A close study of these appearances, therefore, is advisable. The following somewhat brief description of larvae and pupae dead of the disease will suffice in most instances and will serve as a guide to further observations. SYMPTOMS MANIFESTED BY THE BROOD Signs that will determine the exact time of death from American foulbrood have not yet been ascertained. As the larvse and pupae that die of the disease do so at a period in their growth when healthy brood is motionless, lack of motion is no guide. In the descriptions made in the present paper it is assumed that a larva or pupa is dead if it shows a change from a bluish-white, more or less transparent appearance to one that is more nearly white, more nearly opaque, and shows at the same time a change from the normal turgidity to a slightly flaccid condition. The appearance of the larval and pupal remains changes gradually from day to day from that of a healthy brood to that of the dried residue — the scale. A description, therefore, that would be correct for one day would be incorrect probably for the following day. Furthermore, all of these remains do not pass through the same changes. For convenience in description, therefore, the various appearances assumed by them are considered in five more or less arbitrary stages. The interpretation of the descriptions will be aided if the description of healthy brood (p. 3) is borne in mind, since the terms used are similar in both instances. LARV^ DEAD OF AMERICAN FOULBROOD The descriptions made of dead larvae (prepupae) for the most part will be of those that have died during the 2-day quiescent period just preceding the transformation to the pupa, as most larvae dying of the disease succumb at this age. FIRST STAOE The first symptoms of American foulbrood appear about the end of the first week after infection. From the bluish white of the healthy larva the color changes during this first stage of the disease to a very light brown. The conelike anterior third (PI. II, E) having settled somewhat, the apex is now slightly farther from the roof. The surface markings in each of the three thirds (PI. II, H) are very similar to those of a healthy larva. The body wall is easily ruptured, but by care the larva may still be removed intact from the cell. In consistency the decaying tissues are soft and nonviscid. 8 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. Sometimes a considerable portion of a larva is removed piecemeal by the bees during the first stage of decay. The remnant (PI. VI, D), consisting of the posterior third and usually much of the middle third, is found occupying its normal position in the cell. The surface markings are similar to those of the middle and posterior thirds in the first stage of decay. The surface left by the removal of the fragments is somewhat roughened and transverse to the body. These larval remnants are similar to those seen in sacbrood, but in American foulbrood there are fewer of them. When present they form one of the earliest symptoms of the disease. SECOND STAGE By the second week after death the larva shows a slightly deeper shade of brown, although still quite light. The anterior third (PI. II, F) is somewhat darker than the other two thirds. The apex is farther from the roof than in the first stage. The surface markings (PI. II, I) are now less pronounced throughout. An attempt made at this time to remove the dead larva from the cell results usually in rupturing the body wall. The decaying mass can be entirely removed, however. The consistency of the tissue mass in this stage is somewhat similar to that of moist dough. It has not yet reached the true viscid condition, THIRD STAGE By the third week after death the remains have assumed a medium shade of brown, approximating that usually seen in brood cappings. The apex of the anterior third (PL III, D) is now widely separated from the roof of the cell, exposing to view the ventral surface of the remains. The transverse ridges and furrows marking the segments (PI. Ill, G) are practically obliterated. The edges no longer show the deep notches. The side-to-side convexity is decreased. Evidences of developing head and thoracic appendages are sometimes seen. Transverse trachese can usually be observed as white lines across the ventral surface of the abdomen, one in each segment. The body wall is ruptured easily. The decaying mass shows some viscidity and adheres to the walls of the cell. FOURTH STAGE By the fourth week after death the color of the dead larva reaches a deep brown, being in shade similar to that of the average older brood combs. The apex of the anterior third (PI. Ill, E) is raised only slightly above the upper surface of the decaying larval mass. After being imcapped this third is the first to dry, becoming dark and scale- like. Surface markings practically have disappeared; tracheae frequently are still visible, especially in the middle third. The ventral AMERICAN FOULBROOD. 9 surface (PI. Ill, H) is now slightly concave from side to side. The posterior third lies upon the bottom of the cell and extends to the roof. The decaying larval mass is now decidedly viscid in con- sistency. When tested with a toothpick, match, forceps, or any other suitable object it can be drawn out into threadlike strings. This viscid condition of decaying brood is referred to by beekeepers as "ropiness." It is a well known s3rmptom and is much used in the diagnosis of the disease. FIFTH STAGE One month or longer after the death of the larva the remains are found to be a thin mass more or less dry and covering most of the floor, some of the side walls, and the bottom of the cell. The dried mass is known to beekeepers as the ' 'scale." Its color is dark brown, similar to that of old brood comb. That portion which was once the anterior third of the larva bears, if at all, only a slight elevation (PI. Ill, C, F, I; PI. VI, E, H). The ventral surface (PL III, I) is concave from side to side and quite uniform. The posterior third covers the bottom of the cell and extends to the roof. This can be seen especially well by cutting lengthwise the cell with a scale (PI. VI, H) in it. It can be seen also that the thickness of a scale throughout the median line is approximately uniform. As some larvae die of American f oulbrood before the quiescent stage is reached but after being capped, their dead remains may be found occupying not the uniform position just described but various posi- tions in the cell. While the form of the remains, therefore, may vary materially, the color and consistency pass through stages that are similar in all instances. Occasionally death from American foulbrood takes place while the larva is still younger, i. e., before the time of capping has arrived and while it (PI. VI, A) is coiled in the cell. This condition is com- paratively rare. As the form of the remains depends upon the age of the larva at the time of its death, naturally, the form of the remains of a larva dying before the time for capping wiU vary materially from the descriptions above. The color and consistency of such a larva during its decay, however, pass through stages that are similar to those already described. PVPM DEAD OF AMERICAN FOULBROOD Death from American foulbrood does not take place late in the pupal stage but almost always, if not invariably, within the first 2 days after transformation from the larva (prepupa). At this stage the healthy pupss are practically white, with or without pigment in the compound eyes. Pupae dead of the disease resemble in color and consistency larvae dead of the disease. The dead pupae, as in the case of the larvae, are described here in five stages. 10 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. FIRST STAGE During the second week following the infection, symptoms of Amer- ican foulbrood may be seen in pupse. The cell containing a dead pupa is generally found capped at this time. Upon uncapping, the anterior third (PI. IV, B) of the pupa will be recognized as resembling very closely that of a healthy one. The surface markings in general (PI. IV, E) are not particularly unlike those of a healthy pupa. The turgidity is slightly less and the body is to a slight degree more nearly opaque. The body wall is easily ruptured and the tissue mass is soft but not viscid. In general the consistency is similar to that of a larva in the first stage of decay. SECOND STAr.E The process of decay continuing, the tissues soften and the pupal mass settles, distorting its form. The anterior third (PI. IV, C) has settled to the floor of the cell and is separated from the roof by a considerable distance. The face is directed more nearly upward. The surface markings (PI. IV, F) of each third are less distinct. The legs and other appendages rest upon the body. The body wall is at this time very easily ruptured and the tissues are soft. The color and consistency are similar to those of the second stage of decay of the larva. THIRD STAGE During the third stage the appendages are less easily distinguished ; as they settle they merge more or less with the decaying mass of the head, thorax, and abdomen. One exception should be mentioned. Some of the mouthparts (PI. V, A), chiefly the proboscis, adhere to the roof of the cell. This condition has often been noted by the beekeepers. The face of the pupa in this stage is directed still more nearly upward than in the preceding stage. The head is not infre- quently found to be disfigured through drying or bemg gnawed or both. The ventral surface of the middle third (PI. V, D) is still, in general, convex from side to side. The color and consistency are similar to those of the same stage in the larvae. When the pupal mass is removed white lines may be seen in it. These are tracheae. FOURTH STAGE In the fourth stage evidence of drying is marked and pupal re- mains are changed still further. The anterior third (PI. V, B) bears only a slight resemblance to that of the pupa. Through set- tUng of the mass and through drying it is now very much flattened. The ventral surface of the middle and posterior thirds (PL V, E) are still slightly convex from side to side and are roughened, due cliiefly to the remains of the legs and other appendages. The color of the decaying mass in the fourth stage is a dark brown similar to that AMERICAN FOULBROOD. 11 of old brood comb. This mahogany-hued mass is viscid in consis- tency, showing the ropiness that characterizes brood dead of American foulbrood. FIFTH STAGE Finally after a few weeks of decay and drying there is to be found in the cell the more or less dry residue of the dead pupa — the scale (PI. V, C, F; PL VI, F, I). It covers most of the floor and some of the side waUs and only a portion of the bottom. The pupal scale is in many respects similar to the larval scale. It is concave from side to side. The ventral surface is slightly roughened by the dry decayed appendages. The posterior third extends only shghtly upon the bottom of the cell. Tliis can be demonstrated by cuttmg the scale (PI. VI, I) lengthwise as it occupies its position in the cell. The scale is dark brown resembling in shade that of old brood comb. It adheres rather firmly to the cell but with care can be removed from it. When thoroughly dry and removed it is found to be quite brittle. ETIOLOGY PREDISPOSING CAUSES Age. — ^American foulbrood infection takes place only dm-ing the feeding stage of larvse. Death occurs almost invariably after the feeding stage is passed, i. e., after capping, and either while in the larval stage or soon after transformation to the pupa. Older pupae do not die as a result of the disease and adult bees do not become infected. Sex. — ^That worker, drone, and queen larvae are all susceptible to the disease has been demonstrated durmg these studies. Affected drone brood is encountered less often in the diagnosis of this disease than in that of European foulbrood. The writer has encountered queen larvae affected by American foulbrood in experimental colonies only, although very probably diseased queen larvae do occur in nature also. Race. — Thus far no race of bees has been shown to possess com- plete immunity from the disease. In the experimental inoculations recorded in the present paper bees mixed with Italian blood were used for the most part. The queens in many of the colonies were purchased as "untested Italians." At least five colonies of "tested Italians," two of "tested Carniolans," and two "tested Caucasians" were inoculated. Among the colonies used there were also several common black bees. The disease was readUy produced in all of these strains. Furthermore, the results obtained from the examination of numerous samples of diseased brood received from beekeepers through- out the United States indicate that all strains of bees commonly found in American apiaries are susceptible to infection with American foulbrood. No definite conclusion can be drawn at the present time regarding the relative immunity possessed by the different races. 12 BULLETIN 809, U. S. DEPARTMENT OP AGRICULTURE. Climate. — It is quite well known that American foulbrood is very widely distributed. The wi'iter has examined samples of the disease from England, France, Germany, Switzerland, New Zealand, Canada, Cuba, and various parts of the United States. This is sufficient to show that the disease exists mider a great variety of climatic condi- tions. The practical import of the observation is that the presence of American foulbrood m any particular locaUty can not be attrib- uted entirely to the climatic conditions of the region. Season. — ^In general the losses from American foulbrood occur later in the bee season than do losses from either sacbrood or Euro- pean foulbrood. Experimental inoculations have shown that the larvae of bees are susceptible at all seasons and that the disease can be produced whenever brood is being reared. It would seem, there- fore, that the severity of the disease is due more to environmental conditions existmg at the different seasons than to any difference in the susceptibility of the larvse dm*mg these periods. Food. — Since American foulbrood occurs in such widely different localities (see ^' Clinfiate"), wherein the food of bees varies almost as much as it is possible for it to vary, it may be concluded that the quality of the food used by bees has very Httle, if anything, to do in the causation of the disease. Furthermore, it is found experi- mentally that the disease can be produced when the colony is well supplied with food, when there is a moderate quantity present, or when there is a scarcity. It would seem from the facts at hand that the course of the disease probably is governed to some extent indi- rectly by the quantity of food present and to a less degree, if at all, by its quality. The value of these factors has not been determined but it is certainly not great in either case. EXCITING CAUSE During the period from 1885 to 1902 (p. 2), Bacillus alvei was assumed quite generally to be the exciting cause of foulbrood. From 1902 to 1907 much interest was manifested in the problem relating to the cause of the foulbroods. This is shown, by the investigations of Lambotte (13) in Belgium, Burri (6, 7) in Switzerland, Bahr (2) in Denmark, Maassen (14) and Erne (12) in Germany, and those of the wi'iter. The writer's experience with the bee diseases began in 1902 when he was working under the direction of Dr. V. A. Moore. In the first studies made, spores were found in very large numbers in the scales of the ropy foulbrood. As these did not grow on the media commonly used in a laboratory (17) it was at once recognized that they were not the spores of Bacillus alvei. Furthermore, as B. alvei was not found in any of the samples of the ropy disease, the conclusion was drawn very naturally that this bacillus could not possibly be its cause. Bacillus alvei was found to be present, however, with much regu- larity and in great numbers in the brood disease that is not charac- AMERICAN FOULBROOD. 13 terized by ropiness and foul odor — European foulbrood. These observations taught, therefore, that B. alvei was not to be associated with American foulbrood— the ropy, foul-smelling disease — but with European foulbrood, which does not possess these characteristics. After beekeepers had convinced themselves that there were two foulbroods, one with a peculiar foul odor and showing a marked ropiness of the dead brood and the other without these characters, B. alvei unfortunately remained in the literature as the cause of the ropy, foul-smelling disease. This fact accounts for not a little of the confusion that has existed in coiuiection with the brood diseases. It should be remembered that B. alvei is not the exciting cause of any bee disease. The writer found (20) that the spores which had refused to ger- minate on the media ordinarily used in the laboratory would germi- nate and the bacillus would grow in an agar medium in the prepara- tion of which bee larvae were used. Satisfactory cultures for experi- mental purposes were not obtained with this medium, however. Later he (23) succeeded in devising a medium (p. 18) by the use of which pure cultures of Bacillus larvae suitable for experimental work could be obtained in abundance and with such cultures American foulbrood was produced by inoculation, the experimentally produced disease manifesting symptoms that are typical of American foul- brood encountered in nature. BACILLUS LARVAE Bacillus larvae, the cause of American foulbrood, had been seen by microscopists doubtless long before it was cultivated successfully in the laboratory. The species requires special media for its cultivation. Success in the germination of its spores was attained in 1903 on an agar made from bee larvae. Pending more definite information regarding the bacillus the writer (20) referred to it by the term, Bacillus "X." Further knowledge concerning the species was gained during the summer of 1904, and it was given the name Bacillus larvae (Wliite 21, 22). Burri (6) in Swit- zerland, working on the disease entirely independently, also recognized the fact that the spores present in such large numbers in the scales represented a new species that was difficult of cultivation. Maassen (14) has referred to the species as Bacillus brandenburgiensis, and ('owan (10) has referred to it as Bacillus burrii (8). Occurrence. — Bacillus larvae occurs in the brood of bees dead of American foulbrood and where contaminations with such material have taken place. Morphology. — The vegetative form is a slender rod with ends slightly rounded and with a tendency to grow in chains (fig. 1; PI. VII, A). It varies markedly in length, depending largely upon the medium used in its cultivation. On the surface of brood-filtrate agar it is more often from 2.5 to 5 ^ in length and about 0.5 ^ in breadth. In a liquid medium it is usually much longer, becoming then frequently filamentous in character. During the vegetative stage the bacillus undergoes changes (fig. 5; PI. VII, E, F, G, H) seldom noted for the bacteria. The rod possesses numerous flagella which are peritrichic in arrangement (fig. 2; PI. VII, C). Giant ivhips. — Giant whips occur (fig. 6; PI. VII, F, G, H)in large numbers, being found especially numerous in the water of condensation of brood-filtrate agar cultiures. These corkscrewUke structures vary widely in their dimensions from scarcely visible coiled filaments to bodies several micra in diameter. 14 BULLETIN 80!), U. S. DEPARTMENT OF AGRICULTURE. Motility. — The bacillus is moderately motile in young cultures taken from the surface of brood-filtrate agar but is somewhat more sluggish when grown in liquid cultures. Spore formation. — On brood-filtrate agar evidence of spore formation ia present about the third day. The rod is swollen near the center where the spore is formed, being then spindle-shaped (fig. 3; PI. VII, B). In a few days numerous spores free Fig. l.SaciUus lanae: Vegetative form. Fig. 2.— Bacillus larvae: The flagella. from the rods are to be seen (fig. 4; PI. VII, D). They measure about 0.6 by 1.3 /x. In some environments few or no spores are produced. This occurs in liquid media, deep in solid media, and on media containing glycerin, mannite, or glucose. Some of the other sugars and also honey inhibit spore formation. Staining properties. — The rods color readily and uniformly with the analine stains, and are Gram-i^ositive. The spores are quite resistant to stains. Fig. 3. — Bacillus larvae: Spore formation. Fig. 4. — Bacillus larvae: The spores. Oxygen requirements. — ^^'hen bee larvai agar alone is employed and inoculations are made with spores following Liborius' method for anaerobes, growth as a rule has appeared more often near to than on the surface (PI. VIII, E, F, G, H, I). Subcul- tures on brood-filtrate agar (PI. VIII, D) and egg-yolk-suspension agar or their combination yield abundant surface growth. Agar slant. — On the surface of inclined brood-filtrate agar subcultures (PI. VIII, D) grow rapidly, producing a-moderate to heavy growth in 24 hours. This growth tends AMERICAN FOULBROOD. 15 to spread somewhat from the area inoculated, is grayish-white, and slightly viscid. It presents a more or less uniform border, a smooth surface, and a ground-glass appear- ance. Older cultures are less prominent than the younger ones. Agar plates. — Colonies on the surface of agar vary in size depending upon the num- ber present. When only a few are present not infrequently they spread and attain a diameter of a centimeter or more. The border is clearly defined and uniform. The growth is only slightly raised and has a smooth surface and a ground-glass appear- ance (PI. VIII, A, B). Deep colonies vary from lenticular to irregular in form with filamentous outgrowths from portions of their surface (PI. VIII, A, C). Fermentation. — Carbohydrate liquid media as ordinarily prepared are not suitable for the growth of Bacillus larvae. In some of these after a considerable period a slight growth may appear at the bottom of the tubes. A little brood-filtrate or egg-suspension Fig. 5.— Bacillus larvae, illustrating an inter- esting feature of the organism. Fig. 6. — Giant whips from cultures o( Bacillus larvae. added to the media improves it. No visible gas is formed but in some instances slight acidity is produced. Gelatin. — No growth takes place in plain or in brood-filtrate gelatin at tempera- tures at which it remains congealed. Power to resist disinfectants. — The spores of Bacillus larvae are very resistant to heat. When suspended in water the more resistant ones require as much as 100° C. maintained for 11 minutes to destroy them and when suspended in honey require a half hour or more (p. 22). They resist respectively 5 per cent carbolic acid for months; 1 to 1,000 mercuric chlorid for days; 10 per cent formalin for hours; and 20 per cent formalin for minutes; each acting at room temperature. Pathogenesis. — Bacillus larvae is pathogenic for the brood of honeybees. Infection takes place during the feeding period of the larvae by way of the alimentary tract. The brood dies of the disease in the larval, prepupal or early pupal stages. The earliest symptoms to be observed following inoculation usually occur in prepupse during the seventh day after the feeding, rarely earlier. The earliest e\'idence of infection in pupse occurs one or two days later. The bacilli are at this time distributed throughout the body. Adult bees, rabbits, guinea pigs, and rats are not susceptible to infection with the parasite. While the dead brood in the scale stage and in other later stages of decay in American foulbrood invariably contains spores of Bacillus larvae in immense numbers, a microscopic examination of earlier stages shows not only few spores but also a comparatively small 16 BULLETIN" 809, U. S. DEPARTMENT OF AGRICULTURE. number of rods. A partial explanation at least of this rather unex- pected fact was gathered from the study of the organism on artificial media. Here it was found that the rods at times undergo interesting but as yet undetermined changes. On glucose and other media, when the growth is luxuriant and the spore formation is very much inhibited, it is noted that in older cultures rods are comparatively few in number but in their stead are forms which are neither rods nor spores but are small, more or less spherical bodies (PI. VII, F, G, H). The exact transformation that has taken place in the bacillus to produce these bodies is not known. Something of the phenomenon has been learned through a study of the rods stained with a flagellar stain employing in the method a light aqueous suspension of a young culture. In such preparations there is seen within the rods a number of individual elements, apparently, each of which is supplied with flagella. These structures become independent and separate from the rod proper (fig. 5; PI. VII, E, F, G, H). In the cultures are found also, especially in the condensation water of brood-filtrate-agar slant cultures (PL VIII, D), giant whips (fig. 6; PI. VII, F, G, H) of various sizes. That the little-understood changes that take place in the vegetative forms bear a close relation to the formation of the giant whips can readily be suspected. Definite data have been sought to prove the identity of the ropy foulbrood of the different countries. Samples of brood comb con- taining disease material were received from Canada, Cuba, England, France, Germany, New Zealand, and Switzerland. The findings from studies made on these were compared each with the other and with the findings from many samples from the United States. Spores of Bacillus larvae in very large numbers and practically in pure cultures were present in every one. In each instance one of the special media (p. 18, 19) required for the cultivation of this species was needed. Culturally the bacillus was the same from all of the samples. The disease produced by the inoculation of colonies of bees was also similar in every instance. Sera from rabbits im- munized with cultures from American sources agglutinated in high dilutions cultm-es from English, New Zealand, and Switzerland sources at least, these being all that were tested in this way. The evidence, therefore, justifies the conclusion that the ropy foul- brood of each of the different countries mentioned above is identi- cally the same disease. Four rabbits inoculated with pure cultures of Bacillus larvae showed comparatively little reaction. In each instance there was used a moderately clouded suspension in normal salt solution made from the surface of the brood-filtrate-agar slant culture about 24 hours old. Two of them were inoculated subcutaneously with 1 and 2 c. c. of the suspension, respectively; one received the culture intraperi- AMERICAN FOULBROOD. 35 usually spreads more or less rapidly within an infected colony, the fact remains that it frequently does not. It was observed that many diseased colonies could be present in the experimental apiary without causing the mfection to be transmitted to other colonies in the apiary. This fact is naturally very significant. On account of it certain views concerning the manner of transmission of the disease, which might othei'wise be regarded as probable, are rendered untenable. If flowers visited by bees from infected colonies, and later by bees from healthy ones, are a likely source of mfection, or if the water supply is such a source, or if drones are a means by which the disease is likely to be trans- mitted, there woiild have been a different observation to report. In a few instances m the experimental apiary colonies near heavily infected ones became infected. That the disease may have been transmitted through the drifting or accidental straying of bees is one of the possible explanations for the disease in these cases. That a slow form of robbmg might have taken place should be considered also as a possible explanation in these cases. Brood comb contammg brood dead of American foulbrood will transmit the disease when placed in a hive containing a healthy colony. The likelihood that the disease will be transmitted by combs from diseased colonies, which contain honey but no brood, probably is frequently overestimated. It would seem that a spread of the disease in this way would depend considerably upon the amount of infection that was present in the colony from which the combs were removed. This would depend also somewhat upon the presence or absence of brood in the colony to which the combs were given. Sufficient facts are wanting to make definite statements m regard to the probability of infection m such cases. Robbing material con- taining the spores of American foulbrood from any source is likely to transmit the disease although it does not necessarily do so. How often the disease would be transmitted through the medium of hives which have housed infected colonies, if used without flaming, has not been definitely determined. Experimental colonies have been placed in such hives and kept in them for a year with the result that no disease appeared.^ It probably will be found that in many cases the treatment of the hive bodies is not necessary to insure that the disease will not be transmitted by them. From the results obtained by practical apiarists and by observation made in the exper- imental apiary during these studies the fact has been well established, 1 An experiment in which four colonies wereused, wasmade as follows: The insides of the top and bottom board of each of two hive bodies were washed with an aqueous suspension of spore-containLng material and allowed to dry, and the inside of the walls of each of two others were similarly washed and allowed to dry. Colonies were transferred to these four hives during the summer. American foulbrood appeared in the two colonies housed in hives of which the tops and bottoms had been contaminated with disease material but did not appear in the other two. These results, naturally, are not conclusive but they are suggestive. 36 BULLETIN 809, U. S. DEPAKTMENT OF AGRICULTURE. however, that the danger of infection from such hives is enthely removed by carefully flaming them inside. The destruction of the hive is, therefore, never necessary. Chemical disinfectants should not be relied upon for the destruction of the spores of American foulbrood as the spores possess a marked resistance to these agents. If any chemical is used, however, strong solutions of formaldehyde would seem to offer some promise of being efficient. After disease material is removed from the hive by bees and is released from them the chances that a bee will come m contact with the contained spores again is comparatively slight. Furthermore such spores must withstand certam agencies m nature which tend to destroy them, thus decreasing the chances still further that disease will result from them. If exposed to the direct rays of the sun the spores will be destroyed more or less readily (p. 29). Much less is to be expected from fermentation (p. 31) or drying (p. 29). Should the spores reach the soil or running water the likelihood that bees wall come m contact with them and carry them to the feeding brood is indeed slight. Should the spores reach the water supply of the bees and should this supply be a stagnant or slow-moving body of water, the chances naturally would be somewhat increased. Experiments in which queens taken from diseased colonies were introduced into healthy ones were not kept under observation for a sufficiently long period to justify a definite conclusion as to the prob- ability of the disease being transmitted by such queens, but the data secured indicate strongjy that the danger of transmission in this way has been overestimated at times. The favorable results ob- tained by beekeepers using the shakmg treatment should tend to allay fear in regard to the transmission of the disease by this route. Indeed the facts thus far obtained suggest that the transmission of the disease by way of the queen should not be expected. The spread of the disease by means of the clothing or hands of the apiarist is not to be feared especially. The hive tool, if brought in direct contact with dead larvoe in testing for the presence of disease, might serve to transmit infection, but during the usual manipula- tions it would not. The practice sometimes followed by beekeepers of thrusting the hive tool into the soil to clean it would seem to be a safe procedure. Washing the hands with water mstead of disinfec- tants is more convenient and is sufficient. Reasonable care should be taken, of course, that the water used does not become or reach the water supply of bees. DIAGNOSIS American foulbrood is the easiest of the brood diseases to diag- nose. Methods by which this can be done in the laboratory are given in an earlier publication (16). The diagnosis as a rule can be AMERICAN FOULBROOD. 37 made from the symptoms (p. 5) alone. No helpful sign is found in the appearance of the adult bees. A weak colony justifies a suspi- cion that a disease is present. In American foulbrood, as in Euro- pean foulbrood and sacbrood, a sample suitable for diagnosis must contain the remains of brood dead of the disease. The colony sj^mptoms which are usually adequate for a definite diagnosis of American foulbrood are the following: The death of larvae in capped cells after the endwise position has been assumed (Pis. I, II, III, VI) , and the death of pupae soon after transformation (Pis. IV, V, VI), the brown shade of dead brood, the viscidity (ro- piness) of the decaying remains, the character of the scales, and the foulbrood odor. BACTERIOLOGICAL EXAMINATION A conclusivo diagnosis of American foulbrood can always be made from a bacteriological examination of brood dead of the disease. With gross characters suggesting the disease the experienced can frequently make the tentative diagnosis definite by a water mount made from the decaying brood remains, or the scales. A very large nimiber of spores free from rods with no other bacterial species present is the characteristic microscopic picture in the disease. The spores are those of Bacillus larvae (fig. 4; PI. VII, B, D). As a rule, a stained preparation does not furnish much additional aid. In all cases of doubt cultures (agar plates are satisfactory) should be made. The absence of growth on the plates when the other data strongly suggest American foulbrood is sufficient for a positive diagnosis. In routine work the cultures should always be made. Occasionally further evidence that the spores present are those of B. larvae is desired. This can be obtained by the employment of methods and media given in this paper (p. 17). DIFFERENTIAL DIAGNOSIS EUROPEAN FOULBROOD European foulbrood can be recognized (27) in most instances by the death of brood in uncapped cells, the yeUow hue of larvae recently dead changing later often to a brown, and an absence of the foulbrood odor. In cases in which these younger larvae have been removed, the diagnosis of European foulbrood can very often be made from the remains of brood which has died in capped ceUs. In these cases the remains of dead larvae and not of pupae are encountered. These decaying masses pass through a stage at which they are somewhat viscid. The scales are rubberlike ^ in consistency. In these masses and in the scales large niunbers of B. alvei are found. > The term "rubberlike" needs interpretation. When applied to the scales of European foulbrood it means simply that they are less brittle than either those of American foulbrood or sacbrood. The property "elasticity" as usually thought of in common parlance in connection with rubber is not meant. 38 BULLETIN 809^ U. S. DEPARTMENT OF AGRICULTUEE. SACBROOD Sacbrood is recognized (25) by the death of larvae in capped cells and not of pupae, by the sachke appearance of the dead remains, and by the absence of viscidity. The absence of microorganisms in the remains characterizes the microscopic picture. * OTHER CONDITIONS Other brood conditions referred to as chilled brood, overheated brood, starved brood, and in some cases drone brood, must be differ- entiated from American foulbrood. The history of the case, the age of the brood at the time of death, and the absence of ropiness and of the foulbrood odor, usually make the diagnosis comparatively easy. In some of the disorders of adult bees excrement of the adult bees is sometimes found in the cells of the brood combs. When dry the masses resemble somewhat brood-disease scales. The character of the masses together with the microscopic picture is usually sufficient for a diagnosis. In these cases a stained preparation is preferable. The fecal matter contains bacterial rods in large numbers. In all of these conditions the absence of B. larvae furnishes definite evidence that American foulbrood is not present. PROGNOSIS Without treatment the prognosis in American foulbrood is deci- dedly grave, the rule being that the colony sooner or later dies as a result of the disease. This is true for experimental colonies and is true also, as has been proved by the experiences of beekeepers, for colonies in which the disease has occurred through natural means. A colony heavily inoculated experimentally is not destroyed by disease in one month or two months. Loss in strength may be apparent, however, after one month. If a colony is inoculated early in the season and is heavily infected, it may die by midsmnmer; if less heavily infected, it may live longer but die later in the season. The infection may be so shght, indeed, that the colony may not be destroyed directly but be so weakened that it wiU die during the winter or surviving emerge a weakened colony in the spring and then die of the disease during the following bee season. When a colony during the summer contains much dead brood and becomes very weak as a result of the disease, it usually absconds. The queen is to be found among the bees to the last, and stores are not wanting as a rule. A question of considerable interest but one which has not yet been completely answered is: Does an American foulbrood colony ever recover without treatment? Some beekeepers have entertained the AMERICAN FOULBROOD. 39 belief that occasionally it does. Experimental evidence indicates that it probably does. Colonies in the experimental apiary, with a larva or pupa dead of the disease here and there only, certainly did not become badly diseased for more than a year and apparently recovered. To have proved conclusively, however, that such colonies were completely free from possible infection from within would have necessitated observations over a much longer period. It is known that worker, drone, and queen larvae die of the disease. Theoretically it is possible that a colony might become queenless as a direct result of American foulbrood but care must be taken in attributing this condition to the disease, for very often queens are reared in diseased colonies, and to all appearances are healthy. Whether a larva once infected ever recovers from the disease is not known. From the facts at hand it may be concluded, therefore, that the prognosis in American foulbrood in an untreated colony is especially grave. From the practical viewpoint, at least, complete recovery from the disease without treatment, if it occur at all, should be con- sidered for the present to be the exception. While an infected colony may live for a long time and yield a profitable surplus for a considera- ble period, this is not the rule; more often the course of the disease is comparatively short, and the destruction of the colony the outcome. SUMMARY AND CONCLUSIONS A brief summary of facts known about American foulbrood together with a few conclusions drawn from them are given here. Some of the facts have been known for many years, others are of more recent origin, and still others are new. All of them are supported by experi- mental studies recorded in the present paper. 1 . American foulbrood is an infectious disease of the brood of bees caused by Bacillus larvae. 2. All larvae — worker, drone, and queen — are susceptible to the infection; adult bees are not. 3. Man evidently is not susceptible to infection with the organism nor are the experimental animals. 4. So far the disease has not been encountered or produced in other insects than honeybees. 5. The brood of bees can be infected through feeding the spores of the bacillus to a colony. 6. The spores contained in a single scale are more than enough to produce considerable disease in the colony. 7. The portal of entry of the infecting agent is somewhere along the alimentary tract of the larva, most likely the stomach (mid-intestine). 40 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. 8. Practically speaking there are no secondary invaders either during the life of the infected larva or during the decay of the re- mains. 9. The incubation period is approximately 7 days. 10. The brood is susceptible to infection at all seasons of the year. 11. More brood dies of the disease during the second half of the brood-rearing season than during the first half. 12. The disease is present at least in Australia, New Zealand, Denmark, England, Ireland, Germany, France, Switzerland, Canada, Cuba, and the United States. The ropy foulbrood of all these coun- tries is one and the same disease. 13. Occurring as it does in such a wide range of cUmatic condi- tions, it is evident that the presence of the disease can not be attribu- ted alone to any particular climate. 14. The course of the disease in the colony is not afiFected greatly, if at all, by the quality of food used by the bees, or by the quantity present. 15. Colonies in which the disease has been produced through artificial inoculation can be kept in the experimental apiary without transmit- ting the disease to others. This fact is of special importance not only in the technique of making studies, but also in the control of the malady. 16. The spores of American foulbrood remain alive and virulent for years in the dry remains (scales) of larvae and pupae dead of the disease and in cultures that have become and remain dry. 17. The spores are very resistant to most destructive agencies. A variation in resistance is noted both as to the individual spores of a sample and as to the spores contained in different samples. 18. Many of the spores are killed within 1 minute at 100° C. and all of them from some samples are killed in less than 5 minutes. In some instances 96° C. maintained for 10 minutes will destroy all of the spores while 98° C. will often do it. The most resistant of the spores studied when suspended in water have not withstood 100° C. for 11 minutes. 19. The spores withstand more heating when they are suspended in honey or honey diluted with water than when suspended in water. 20. The spores suspended in honey or diluted honey can be de- stroyed by 100° C. but it may require half an hour or more to do so. 21. American foulbrood spores when dry were destroyed by the direct rays of the sun in from 28 to 41 hours. 22. The spores when suspended in honey and exposed to the direct rays of the sun were destroyed in from 4 to 6 weeks. 23. The spores when suspended in honey and shielded from direct sunlight remained alive and virulent for more than a year. It AMERICAN FOULBROOD. 41 is very likely that they are capable of remaining so for a very much longer period. 24. The spores resisted the destructive effects of fermentation for more thaii 7 weeks at incubator and outdoor temperatures, respect- ively, and probably are able to withstand these agencies for a very much longer period. 25. The spores resist carbolic acid at room temperature in strengths ordinarily used as a disinfectant for periods of months; 1 to 1000 mercuric chlorid for days; 10 per cent formalin for hours. 26. Experiments recorded in the present paper indicate that drugs do not materially affect the course of the disease. 27. American foulbrood infection is transmitted primarily through the food of bees; possibly at times to some extent through their water supply. Robbing from the diseased colonies of the apiary, or from neighboring apiaries, is the most likely mode by which the disease is transmitted in nature. 28. The placing of brood combs containing diseased brood with healthy colonies will result in the transmission of the disease. 29. Flowers should not be considered as a likely medium through which infection may take place. 30. Whether the disease is ever transmitted by queens or drones has not been determined. That they have been overestimated at times as possible sources of infection seems likely. 31. It is quite probable that in many cases hives which have housed colonies infected with American foulbrood will not transmit the disease to healthy colonies transferred to them. Results from the present studies confirm the observation made by beekeepers that danger from this source may be removed by properly flaming such hives inside. 32. The clothing of those about an apiary, and the hands of the apiarist are not fruitful sources for the transmission of the disease. 33. Tools and bee supplies generally about an infected apiary wUl not transmit the infection in the absence of robbing from those sources. 34. American foulbrood usually can be diagnosed from the symp- toms alone. A definite diagnosis can always be made from suitable samples by bacteriological methods. 35. The prognosis in the disease in the absence of treatment is decidedly grave but with proper treatment it is favorable. 36. From the technical viewpoint many of the problems consid- ered in these studies have been solved only partially; from the prac- tical point of view, however, the results are sufficient to make a logical, efficient, and economic treatment of American foulbrood possible. 42 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. LITERATURE CITED 1. Bahr, Louis. 1904. Vore bieygdomme. Foredrag holdt ved DBF's Diskussionsm<^de i Grejs- dalen den 11. Septbr. 1904. (Efter Foredragsholderens manuskript). Roskilde [1904]. 17 p. (Saertryk af "Tidsskrift for Biavl" Nr. 16 og 17, 1904). 1906. Om Aarsagen til Bipesten og dennes Beksempelse. Foredrag af Drwlsege L. Bahr holdt ved DBF's Diskussionsm^de d. 2 Septbr. 1906 i Esbjerg, 13 p. Ssertryk af "Tidsskrift for Biavl," Nr. 17. 1915. Sygdomme hos Honningbien og dens Yngel. Meddelelser fra den KgL Veterinaer.-og Landboh(^jskole8 Serumlaboratorium, XXXVII. 109 p.» 11 fig. 1916. Die Krankheiten der Honigbiene und ihrer Brut. Deutschen Tierarz- tlichen Wochenschrift, 24 Jahrgang, Nr. 28 u. 29. 19 p., 2 fig. 5. Beuhne, F. R. 1913. Diseases of bees. Foulbrood. In Journal of the Dept. of Agr. of Vic- toria (Australia), v. 11, pt. 6, p. 367-371, 2 fig. June. 6. Burri, R. 1904. Bacteriologische Forschungen uber die Faulbrut. In Schweizerische Bienenzeitung, No. 10, p. 335-342, Oct., and No. 11, p. 360-365, Nov. 1906. Bacteriologische Untersuchungen uber die Faulbrut und Sauerbrut der Bienen. 39 p., 1 pi. Vorwort von U. Kramer. Jan. 1917. Der gegenwartige Stand der Faulbrutforschung. 20 p. In Schweizerischen Bienenzeitung, Jahrgang 1917, v. 40, no. 1, January. 9. Cheshire, F. R., and Cheyne, W. W. 1885. The pathogenic history and history under cultivation of a new bacillus (B. alvei), the cause of a disease of the hive bee hitherto known as foul- brood. In Jour. Roy. Micros. Soc, Ser. II, v. 5, pt. 2, p. 581-601, pi. X-XI, August. 10. Cowan, T. W. 1911. British bee-keeper's guide book ... Ed. 20. viii, 226 p.. 143 fig. London. 11. Dzierzon, Johann. 1882. Dzierzon's rational bee-keeping . . . Tr. by H. Dieck and S. Stutterd. Ed. and rev. by Charles Nash Abbott . . . xvi, 350 p., illus. London. 12. Erne, Dr. Fritz. 1906. Bakteriologische Untersuchungen uber die Faulbrut und die Sauerbrut der Bienen. In Die Eiu-opaischen Bienenzucht, p. 148-151, November. 13. Lambotte, Ul. 1902. Recherches sur le microbe de la "loque," maladie des abeilles. Travail du laboratorie de I'institut de pathologic et de bacteriologie de I'uni- versit6 de Li^ge. In Annales de I'institut Pasteur, v. 16, no. 9, p. 694-704, September 25. 14. Maassen, Albert. 1906. Faulbrutseuche der Bienen. Mitteilungen aus der kaiserlichen biolo- gischen Anstalt fiir Land-und Forstwirtschaft, Heft 2, p. 28-29. June. AMERICAN FOULBROOD. 43 15. Maassen, Albert. 1908. tJber die unter den Namen "Faulbrut" bekannten seuchenhaften Bru- terkrankungen der Honigbiene. Mitteilungen aus der kaiserlichen koniglichen Anstalt fur Land-und Forstwirtschaft, Heft 7, 24 p., 4 pi. September. 16. McCrat, a. H., and White, G. F. 1918. The diagnosis of bee diseases by laboratory methods. U. S. Dept. Agr. Bui. no. 671, 15 p., 2 pi. June 21. 17. Moore, Veranus, and White, G. F. 1903. A preliminary investigation into the cause of the infectious bee dis- eases prevailing in the state of New York. N. Y. (State) Dept. Agr. 10th Ann. Kept. Com. Agr. for 1902, p. 255-260, 2 pi. January 15. 18. Muck, Oswald. 1915. Seuchen der Bienenbrut. In Wiener tierarztlichen Monatsschrift, Jahrg. 11, Heft. 3, p. 124-139, 6 fig., 2 pi. 19. SCHIRACH, A. G. 1771. Histoire naturelle de la reine des abeilles, avec I'art de former des essaims. Ixiii, 269 p., 3 pi. La Haye. 20. White, G. F. 1904. The further investigation of the diseases affecting the apiaries in the State of New York. New York (State) Dept. Agr. 11th Ann. Rept. Com. Agr. for 1903, p. 103-114, January 15. 21. 1905. The bacterial flora of the apiary with special reference to bee diseases. Thesis. Cornell University, Ithaca, N. Y. 22. 23. 24. 25. 26. 27. 28. 1906. The bacteria of the apiary, with special reference to bee diseases. U. S. Dept. Agr. Biu-. Ent. Tech. Ser. no. 14. -50 p. 1907. The cause of American foulbrood. U. S. Dept. Agr. Bur. Ent. Giro. 94. 4 p. July 29. 1914. Destruction of germs of infectious bee diseases by heating. U. S. Dept. Agr. Bui. 92. 8 p. May 15. (See p. 4.) 1917. Sacbrood. U. S. Dept. Agr. Bui. 431. .55 p., 33 fig., 4 pi. February 9. (Professional Paper.) 1919. Nosema-disease. U. S. Dept. Agr. Bui. 780. .59 p., 7 fig., 4 pi. June 12, (Professional Paper.) 1920. European foulbrood. U. S. Dept. Agr. Bui. 810. 39 p., 6 fig., 8 pi. (In press.) 1919. Unheated egg-yolk media. In Science. N. S. v. 49, no. 1267, p. 362, ApriMl. 29. Zander, Enoch. 1910. Die Faulbrut and ihrer Bekampfung. 32 p., 8 fig., 4 pi. Stuttgart. (Hand- buch der Bienenkiinde I.) EXPLANATION OF PLATES PLATE I Brood aflected with American foulbrood: A. — American foulbrood produced by the feeding of pure cultures of Bacillus larvae, showing the later stages of decay of the brood. B.— American foulbrood produced experimentally, showing irregular distribution ofeapped and un- capped cells, punctured caps, and scales. C. — American foulbrood, showing sunken and punctured caps. The sample was shipped by mail for diagnosis. PLATE II Healthy bee larvse (prepupse) and first and second postmortem stages of American foulbrood larvae as described in the present paper: A. — Cap of cell containing a healthy larva. Being recently constructed, it is convex. B. — Cap of cell containing a larva recently dead of the disease. It is not unlike many of the caps over healthy larvae. C— Cap of cell containing a diseased larva. It is slightly sunken. D.— End view of a healthy larva. The cap was removed artificially. (In thelaboratory this is done con- veniently with fine pointed cur\'ed forceps or a dissecting needle.) E.— End view of larva illustrating the earliest (first) stage of American foulbrood. F.— End view of larva which has reached the second stage. G.— Ventral view of a healthy larva of the age at which death from American foulbrood frequently occurs. H.— Ventral view of larva illustrating the first stage after death. I. — Ventral view of larva showing second stage. PLATE III Third, fourth, and fifth stages of decay in American foulbrood larvae: A. — Punctured cap of cell containing a dead larva. B.— End view of larva illustrating the fourth stage of decay. The cap was removed artificially as indi- cated by the torn edges. C— End -piew of larva illustrating the fifth stage of decay, the scale. The cap had been removed by the bees, as shown by the smooth condition of the mouth of the cell. D.— End view of larva, third stage. E. — End view of larva, fourth stage. F. — End view of larva, fifth stage — the scale. G.— Ventral view of larva, third stage. H. — Ventral view of larva, fourth stage. I. — Ventral view of larva, fifth stage — the scale. PLATE IV Healthy pupae and first and second postmortem stages of American foulbrood pupae as described in the present paper: A. — Healthy pupa of the age at which death from American foulbrood occurs. B. — End view of pupa, first stage of the disease. C— End view of pupa, second stage. D. — ^Ventral view of pupa of the age at which death from American foulbrood occmrs. E. — Ventral view of pupa, first stage. F. — Ventral view of pupa, second stage. PLATE V Third, fourth, and fifth postmortem stages of American foulbrood as described in the present paper: A. — End view of pupa, third stage. B. — End view of pupa, fourth stage. C. — End view of pupa, fifth stage— the scale. D. — Ventral view of pupa, third stage. E.— Ventral view of pupa, fourth stage. F.— Ventral view of pupa, fifth stage— the scale. 45 46 BULLETIN 809, V. S. DEPARTMENT OF AGRICULTURE. PLATE VI Various phases of the appearance of brood dead of American foulbrood: A.— Larva in cell not yet capped, dead of American foulbrood. B.— A developing bee dead of American foulbrood immediately before entering the pupal stage. C. — A punctured cap with two holes. D.— Larva dead of American foulbrood gnawed and partially removed by the adult bees. E.— End view of larval scale. The cap was removed artificially. F.— End view of pupal scale. G.— Lateral view of healthy larva in capped cell. H.— Lateral view of larval scale in the cell cut lengthwise. The cap had been removed by the adult bees. I.— Lateral view of pupal scale in positioft in cell cut lengthwise. PLATE VII Vegetative form, spore formation, spores, flagella, and giant whips of Bacillus larvae. Photomicrographs: A.— Vegetative form from a 24-hour culture on the surface of brood-filtrate agar. X 1000. B._Spore formation and sporesfrom the siu-face of brood-filtrate agar. X 800. C— Flagella. Stained by Loefiier's method as modified by Johnson and Mack. X 1000. (Retouched.) D.— Spores from a stained smear made directly from a decaying larva dead of American foulbrood. X 1000. E.— The production \^ithin the rods of separate elements, each being supplied i^-ith flagella. X 1000. F.— Giant whips and small spherical bodies in the older cultures of B. larvae obtained from the water of condensation of a brood-filtrate-agar slant culture. Fixed in 50 per cent formalin and stained with carbol fuchsin, X 1000. G.— Small and larger giant whips, and spherical bodies similar to those of F. X 1000. A similarity existing between the microscopic appearance of giant whips and spirochetes caused Maassen (14) to believe them to be spirochetes and to introduce the name Spirochaeta apis. H.— Very large whip, and spherical bodies similar to those of F and G. X 1000. PLATE VIII Colonies and cultures of Bacillus larvae: A.— Surface colony and deep colonies as seen in brood-filtrate agar plates. B.— Surface colony on brood-filtrate agar more highly magnified. C. — Deep colony in the same medium, magnified. D.— Surface growth on brood-filtrate-agar slant. E.— Culture of Bacillus larvae in bee-larvse agar when the Liborius method is used. The growth is throughout the upper third of the medium. F.— The culture is similar to E but the growth is nearer the surface. G.— The culture is restricted in the agar and is near to but not at the surface of the medium. H.— The growth is less restricted in area and farther from the surface than in G. I.— The growth is very restricted in area and is below the middle of the medium. o Bui. 809, U. S. Dept. of Agriculture. Plate I. American Foulbrood. Bui. 809, U. S. Dept. of Agriculture. Plate II. American Foulbrood- Bui. 809, U. S. Dept. of Agriculture, Plate 111. American Foulbrood. Bui. 809, U. S. Dept. of Agriculture. Plate IV. ,0 E American Foulbrood. Bui. 809, U. S. Dept. of Agriculture. PLATE V. American Foulbrood. Bui. 809, U. S. Dept. of Agriculture. PLATE V I . American Foulbrood- Bui. 809, U. S. Dept. of Agriculture. PLATE VII. American Foulbrood. Bui. 809, U. S. Dept. of Agriculture. PLATE VI American Foulbrood. AMERICAN FOULBROOD. 17 toneally and one intravenously. The temperature taken of the latter two throughout the day of the inoculation was found to be increased about 3° F. following the inoculation. The next day they were somewhat sluggish with impaired appetite. The temperature, how- ever, soon declined to normal and their feeding became normal. The 1,500 gram rabbits w^ere from 100 to 200 grams light in weight on the day after the inoculation but soon the loss was regained. Observations were made over a period of 2 months. Autopsies on the chloroformed animals showed no abnormalities worthy of note. Four guinea pigs inoculated subcutaneously with cultures similar to those used in the rabbit inoculations showed only slight reaction. The temperature of the one taken tliroughout the day of inoculation showed a rise of about 2° F. but was normal thereafter. No loss in the weight of these 400 to 500 gram guinea pigs was appreciable. One died in 10 days, apparently from causes foreign to the inocu- lation. The remaining three were chloroformed after 6 weeks and presented no abnormalities of jiote at autopsy. Two gray rats were inoculated subcutaneously, one with the vege- tative and the other with the spore form of Bacillus larvae. These proved to be refractory. After 4 weeks they were chloroformed and at autopsy no abnormality of interest was noted except a slight infiltration at the point of inoculation in the animal receiving the vegetative culture, and a small abscess in the case of the animal receiving the spore suspension. In the abscess spores were present. TECHNIQUE Much of the time devoted to the study of American foulbrood has been consumed in the study of technique. Special media were re- quired and the method of conducting the experimental studies had to be developed. MEDIA Failure to recognize the fact that the spores occurring in such large numbers in the brood dead of American foulbrood do not grow on the media ordinarily used in the laboratory has caused a number of workers to go astray and has contributed not a little to the confusion that has existed concerning the brood diseases. Lambotte (13) in Belgium experienced some difficulty in obtaining cultures following inoculations with the spore-bearing disease material. He made a medium using the brood of bees in its preparation and with it obtained a gi'owth which he interpreted as being the species represented by the numerous spores present in the brood dead of the disease. The interpretation was evidently incorrect. The culture which he ob- tained he identified as a member of the mesentericus group. In 1903, before learning of Lambotte's work, the writer (20) had made an agar using bee larvae and obtained a germination of the spores of the 132862°— 19— Bull. 809 2 18 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTUEE. ropy foulbrood. In this bee-larvaa agar^ the growth is always slow a^d somewhat feeble. Spores are produced to a slight extent only or not at all. Such cultures are unsuitable for experimental purposes. A medium containing a sterile filtrate ob- tained from the brood of bees was devised by the writer (23) which meets the requirements of experimental studies. It is prepared as follows: Healthy larvae taken from the brood comb are crushed and the crushed mass is diluted with water to several times its volume. This is placed in a flask, and after adding a few CO. of chloroform the flask is stoppered and allowed to remain at incubator temper- ature overnight. The suspension then can be Pia. 7. — A convenient method for obtaining sterile filtrates. filtered easily by using any bacteria-proof filter^ (fig. 7). The filtrate is pipetted aseptically into sterile tubes and stored until 1 Larvse, prepupje, or pupae recently transformed furnish satisfactory brood material for making the bee-larvae media. These are picked from the brood combs, crushed, and used as meat in formulas ordinarily followed in making bouillon and agar. Excess heating should be avoided. In making the inoculations an aqueous suspension of the spore-containing material heated to 100° C. for one or two minutes should be used. A tube of the special agar liquefied is inoculated with a loopful of the heated suspension. After cooling it is incubated at about 38° C. Three days or more may be required to produce a visible growth (Pl.VIII, E, F, G,H,I). *In caring for and using the filter cylinders in these studies the folio wlngcourse has been pursued: The cyUndcris immersed in water for a few hours or over night and is then sterilized in the autoclave. After being used for filtering a brood suspension it is washed in water without scrubbing. To remove f urtherthe brood material water is drawn through the cylinder under pressure as in filtering. Asaruleit is again im- mersed in water and allowed to remain over night, after which, the filter being assembled, water is filtered again. Itis then ready for use and may be stored until needed. Cylinders which have been used repeat- edly in this way have lost none of their efficiency. The vacuum chamber is made in two sections similar toaNovyjar. Vaseline is used on the wide ground flanges. No rubber band or clamps are needed. The flasks usedin the vacuum chamber of the apparatus before being sterilized are stoppered withcotton wrapped about a small-sized test tube. The cylinder having born immersed in water is sterilized together with the rubber connections and glass tubing in place, but without the glass mantle. In assembling the apparatussterilized connection between the cylinder and flask in the vacuum chamber is readily ob- tained through the opening in the plug left by removing the test tube. After a vacuum has been estab- lished in the chamber the stopcock may be closed and the pump turned off. AMERICAN FOULBROOD. 19 needed. About 1 c.c. of the brood filtrate is added to each 5 c.c. of agar hquefied and cooled to about 50° C. After being inclined the medium is ready for use and may be stored. Plating for pure cul- tures may be done by adding the brood filtrate to the agar at the time the plates are made.^ Wlien it is desired to obtain cultures from the spore form of Bacillus larvae, brood-filtrate agar is not a suitable medium in itself. The spores should be germinated first by inoculating bee-larvae agar (p. 18) or egg-yolk-suspension agar. By plating in 2 or 3 days the vegetative form thus obtamed, using brood-filtrate agar, pure cultures can be assured. Studies have shown, however, that Bacillus larvae is present in practically pure cultures in brood dead of the disease, and by heatmg the spore-containing material in aqueous suspension at 100° C. for one or two minutes as suggested above the occasional contaminating organisms are eliminated usually without plating. Bee-larvaB agar is limited in its usefulness on account of the large amomit of brood required in its preparation. A more suitable me- dium, therefore, was sought. An unheated egg-yolk agar was found by the writer (28) to be very satisfactory. This is prepared as follows: Fresh eggs are used. After havmg been immersed in a dis- infectmg solution the shell is broken, the white of the egg is poured off, and the yolk is dropped into a flask contaming about 70 c.c. of sterile water. With a sterile pipette the aqueous suspension of yolk, resulting from agitating the flask, is transferred to sterile tubes and stored until needed. In maldng an egg-yolk-suspension agar about 1 c.c. of the egg-yolk suspension is added to each 5 c.c. of agar in tubes liquefied and cooled to about 50° C. The agar is inclined and may be stored until needed.^ 1 Larvae, prepupae, or young pupse may be used in obtaining the brood filtrate. The use of sterile water is a good precaution in making the dilution of the crushed brood mass. A dilution of 1 to 10 up to 1 to 50 gives very satisfactory results, although lower and higher dilutions have been used with success, while at incubator temperature an autodigestionapparenty takes place. The "autodigested" suspension is passed through filter paper before the bacteria-proof filter is used. For the latter the Pasteur-Chamber- land F is very satisfactory. (Fig. 7.) The B grade has been used but the filtering is slower. The Bcrkefcld N has also been employed, but is less efficient. In using this latter type of filter gravity alone should be employed. If the weather is warm, or if the filtering is not to be done for a few days, the aqueous suspension may be allowed to remain at room temperature. This permits the changes which take place to advance sufficiently to make the filtering process comparatively easy. The chloroform saturated sus- pension of brood material has been kept at room temperature for more than three years without its useful- ness being impaired to any appreciable extent. After a considerable period the suspension becomes practically sterile and has been used without being filtered. When an early use of the filtrate is desired its sterility should be tested by placing the tubes containing it at incubator temperature for a few days. Sometimes it is desirable to use other brood-filtrate media than the agar. In this event a small amount of brood-filtrate may be added to any one of the media ordinarily used in the laboratory. Such special media usually support a g^o^^■th of Bacillus larvae when inoculated \\ith the vegetative form of the organism. ^Kggs obtained from the market labeled "strictly fresh" have been suitable as a rule for the egg-yolk suspension. Eg js knoATi to have been recently produced are to be preferred, however. Almost any of the more common e Tic'ent dsinfecting solutions may be employed for sterilizing the eggshell. Aqueous solution of mercuric chlorid 1:1000; carbolic acid 5 per cent; formalin 10 per cent; and alcohol each have boon used, mercuric chlorid being preferred. To insure sterility autoclaving of the flasks with the con- tained water has been practiced. A wide-mouthed flask is preferred. In securing the yolk aseptlcally, it 20 BULLETIN" 809, U. S. DEPARTMENT OF AGRICULTURE. EXPERIMENTAL INOCULATIONS A very satisfactory colony for inoculation purposes is a queen- right nucleus that can be accommodated comfortably on from 3 to 5 brood frames approximating the Langstroth size. The hive, the arrangement of the apiary, the colony, the feeding, and the manipu- lations in general are similar in the experimental study of American foulbrood to those employed by the writer in the study of sacbrood (25), Nosema-disease (26), and European foulbrood (27). The inocu- lated colonies were, therefore, in the open. Precautions tending to minimize the likelihood of robbing, swarming, absconding, and the accidental straying or drifting of bees should not be overlooked. The inoculations are made by feeding a suspension of the spores of Bacillus larvae in sirup or honey, the spores being obtained from pure cultures or from brood dead of American foulbrood. Three or four tubes of spore-containing cultures grown on brood-filtrate or egg- yolk-suspension agar furnish a suitable quantity of the virus. Like- wise the spores contained in three or four scales give satisfactory results. In some experiments, feedings made on successive da3^s are desirable. Direct inoculation by means of a capillary pipette is less satisfactory in experiments on American foulbrood than in those on sacbrood.^ The first symptoms of American foulbrood to be observed following the inoculation usually appear in the worker larvae by the end of the first week and not earlier than the sixth day. The affected larvss are in capped cells. Not infrequently the first evidence of disease noticed is the remains in cells here and there of partially removed is convenient to break the sterilized shell about one end of the egg, remove the pieces with flamed forceps, and, after making a small hole in the other end, pour off the white. The yolk remaining is poured into the flask. By breaking the limiting memljrane of the yolk at the time of pouring into the flask the process is facilitated. Different degrees of dilution of the egg yolk have been employed, 60 c.c. to SO c.c. of watergiv- ing satisfactory results. Any ingredient, such as the sugars, desired in the final medium may be added to the water of the flasks. The egg of the hen has been used in most of the studies reported here. Eggs of the duck were also found to be suitable. Occasionally a contamination of the egg suspension will be found. In these cases the suspension becomes firm after incubation resembling then the consistency of coagulated milk cultures. If the contamination is present it will be apparent in a few days after storing, or it maybe; determined earlier at incubator temperature. The sterile egg-yolk suspension as an ingredient for special media retains its efficiency for a long period. Indeed an aqueous suspension of the dry residue from the yolk supcusion was found to be efficient after 3 years, ahhough its value is then apparently somewhat impaired. Maa.sscn (15) used bee-larvae and also brain agar in studying Bacillus larvae. The egg-yolk suspension agar, in the writer's experience, has advantages over the brain agar. 1 Cultures on the surface of agar are readily suspended in water as is also the disease material in the decay- ing larvffi and scales of American foulbrood. These aqueous suspensions arc used in making the sirup Oi' honey suspensions. From 1 to 3 c.c. of water for each scale is a convenient proportion. Sirup is made by bringing to the boiling point an aqueous suspension of granulated sugar in which the sugar used exceeds that of the water. When honey is used for the spore suspension it is diluted with water, the amount of water added being slightly less than that of honey. In most instances honey is less suitable than sirup Inasmuch as it tends to encourage robbing especially during a dearth of nectar, it needs to be sterilized before use, and is more expensive. It was found that less than one scale is sufficient disease material to producea considerable amount of disease in the colony. In some experiments one scale, therefore, might supply all of the spores needed although the use of a somewhat greater quantity of material is advisable in most inslauces. AMERIOAlSr FOULBROOD. 21 larvae (PI. VI, D), about one- third of the larva bemg removed. The transparency seen in healthy larvae disappears, and the bodies of the ones affected become more nearly opaque. Soon the color assumes a tint of brown which deepens as the process of decay continues, pass- ing through chocolate, coffee, and mahogany shades. In the earliest stages of the disease the body wall is ruptured more easily and the tissues are softer than in healthy brood. As the disease advances a ropiness of the decaying brood is to be observed, the characteristic odor appears, and a loss in the strength of the colony is evident. After a time an irregularity in the appearance of the brood comb is to be seen, the capped and uncapped brood being abnormally distributed with here and there perforated and sometimes sunken caps (PI. II, A, B, C). Still later the scales are found. The time at which the various symptoms appear varies. Climatic conditions, the amount of infection present, and the initial strength of the colony are some of the more important factors causing the variations. Whether a hive which has housed a colony infected with American foulbrood will transmit the disease is a question which has not been altogether solved but in experimental work all hives that have housed such colonies should be disinfected before they are used again. This can be done satisfactorily by flaming the inside of the hive. If gas is available the Bunsen burner is very satisfactory for this purpose. That no fear need be entertained from hives which have been flamed properly is demonstrated by the fact that the disease has not been transmitted by flamed hives in the experiments made in the present studies, although the number of such hives which have been used is large. The length of time that a sirup or honey suspension of Ameri- can foulbrood spores stands before the inoculation of a colony is made need not be considered in experimental work. From what is known of American foulbrood its transmission by queens is not to be expected. Two queens from diseased colonies were introduced into healthy ones in October and the colonies were kept under observation until the following July. The disease did not appear in either of them. In a number of instances during the investigations queens from American foulbrood colonies were used to queen nuclei made by division of healthy colonies. These colonies were under observation for weeks or months before they were used for American foulbrood experiments. In no instance was the disease observed to result from the use of these queens. The conclusion is reached, therefore, that for most experimental work at least the possi- ble previous relation of the queen to diseased colonies need not concern one. The combs from an American foulbrood colony never should be given to a colony which is to be used for experimental purposes. 22 BULLETIN" 809, TJ. S. DEPARTMENT OF AGRICULTURE. The bees from diseased colonies may be used again in certain well- selected cases by transferring them to another hive which may be supplied with comb from healthy colonies or with foundation strips or full sheets. To insure that no infection is present a colony treated in this way should be under observation for a con- siderable period after brood rearing has begun in it before it is used again. Further reference to the technique used in these studies will be made as the experiments are recorded. THERMAL DEATH POINT OF AMERICAN FOULBROOD SPORES In an earlier paper (24) the results from preliminary experiments were given which indicate the amount of heating that is necessary to destroy the spores of American foulbrood when they are suspended in water. The importance of heat as a means for the destruction of these spores in practical apiculture is so great that a further development of the subject dm-ing these investigations seemed justifiable. RESISTANCE OF AMERICAN FOULBROOD SUSPENDED IN WATER TO HEATING In preparing the spore material for heating a much diluted aqueous suspension of the disease material is drawn into capsules (fig. 8) made from glass tubing of small bore. After being sealed in a flame they are immersed in a water bath having a temperature and for the period desired in the heating.^ Cultures then are made using a loopful of the suspension from the capsule and brood-filtrate-egg-yolk-suspension agar. The results of the experiments given in the following todetermMng^the table indicate the approximate amount of heating that thermal death jg necessary for the destruction of the spores suspended in water. point of spores. Tablr I. — Preliminary experiments indicating the thermal death point of the spores oj Bacillus larvae suspended in water ^ First Set of Experiments. Disease Materia! Received from Americ« Temperature. Period of heating. Besults as shown by cultures, October, 1913. °C. "F. Minutes. 0 0 0 Numerous spores aUve (check). 90 194 20 Many spores not killed. 94 201 10 One spore not killed. 96 205 10 All spor«e killed. 97 207 10 98 208 10 Do! 99 210 10 Do. 100 212 5 Do. 100 212 10 Do. I 1 About two minutes are allowed for the suspension within the capsule to reach the temperature of tbei water outside before time is reckoned. • 2 Fractions are omitted in this paper, the nearest whole number being given. AMERICAN FOULBROOD. Second Set of Experiments. Disease Material Received prom England. 23 Temperature. Period of heating. Results as shown by cultures, October, 1913. "C. 0 91 9.5 96 98 99 100 100 100 100 100 "F. 0 196 203 205 208 210 212 212 212 212 212 Minutes. 10 10 10 10 10 10 1 2 3 4 5 Numerous spores alive (check). Spores not killed, about fy as many as in check. All but 2 spores killed. All spores killed. Do! Spores not killed, about i as many as in check. All but 100 spores killed. All but 20 spores killed. Do. Do. Third Set of Experiments. Disease Material Received from France. Temperature. Period of heating. °C. °F. Minutes. 0 0 10 90 194 10 92 198 10 93 199 10 94 201 10 90 205 10 98 208 10 99 210 10 Results as shown by cultures, October, 1913. Numerous spores alive (cheek). Spores not killed, almost as many as in check. Spores not killed, about J as manv as in check. All but 30 spores killed. All but 100 spores killed. All but 12 spores killed. All hut 1 spore killed. All spores killed. FotTRTH Set of Experiments. Disease Material Received from Cuba. Temperature. Period of heating. Results as shown by cultures, .January, 1915. "C. "F. Minutes. 92 198 10 Numerous spores not killed. 93 199 10 Do. 94 201 10 Fewer spores not killed. 95 203 10 Do. 96 205 10 Do. 97 207 10 Fewer alive than at 90° C. 98 208 10 Fewer alive than at 97° 0. 99 210 10 About 12,000 spores not killed. 100 212 10 About 20O spores not killed. 100 212 11 Do. 100 212 12 Do. From Table I it will be observed that all the spores taken from the American sample were killed by heating at 96° C. for 10 minutes; that all of those from the English sample were also killed at 96° C. in 10 minutes; that all of those from the French sample were killed at 99° C. in 10 minutes; and that all of those from the Cuban sample were killed at 100° C. in 11 minutes. Spores from the different samples varied, therefore, in their resistance to heating, those from the American and Enghsh samples being slightly less resistant than those from the French one which showed in turn slightly less resist- ance than those from the Cuban sample. By a comparison of the results it will be observed that many spores among those heated were destroyed at 90° C. in 10 minutes and that the percentage remaining alive when the higher temperatures were used rapidly decreased. The results indicate also that by increasing the temperature the time required to destroy the spores is decreased and vice versa. I! 24 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. Further studies relative to the thermal death point of the spores were made on disease samples received from different localities in the United States. The technique followed is similar to that given on page 22. The results obtained are summarized in Table II: Table II. ■Variation in thermal death point of American foulbrood spores from differ- ent localities in the United States ' Period of heating. Source of samples. Temperature. Wash. Minn. Nebr. Ohio. III. Colo. Wis. Pemi. 4529. + + + Perm. 4507. Ohio 4519. Mont. "C. 100 100 100 98 95 "F. 212 212 212 208 203 Minutes. 10 5 1 10 10 - - + + + + + + + + + + + + + + H- ■f 1 The minus sign in Tables II and IV indicates that the spores were not all killed and the plus sign that all of them were killed. The results in Table II show that in none of the 1 1 cases were all of the spores killed in 1 minute at 100° C; in 5 of the cases they were all killed in 5 minutes at this temperature while in 6 of them they were not. In two instances out of the 11 they were not killed in 10 minutes at 100° C. They were killed, however, at this tempera- ture in 11 minutes. It will be seen also that in no instance were all of the spores killed at 95° C. in 10 minutes, but in 5 of the 11 cases all of them were killed at 98° C. within the 10 minutes. By these results it is shown again that the spores of different samples do vary in their resistance to heat. No conclusion is drawn regarding the cause, environmental or otherwise, for this variation. TIME A FACTOR IN THE DESTRUCTION OF AMERICAN FOULBROOD SPORES BY HEAT Suspensions of scales of American foulbrood in water were heated for different periods at 100° C. In the following table are recorded the results obtained when the spores from the Cuban sample were used : Table III. — Period of heating American foulbrood .spores a factor when 100^ C. is used Temperature. Period of heating. °C. °F. Minutes. 100 212 0 100 212 1 100 212 5 100 212 6 100 212 7 100 212 8 100 212 9 100 212 10 100 212 11 100 212 12 100 212 13 100 212 14 100 212 15 Results as shown by cultures, February, 1915. 25,000 spores present (estimated). 4,000 spores not killed (estimated). 148 spores not killed. 220 spores not killed. 248 spores not killed. 44 spores not killed. 7 spores not killed. 14 spores not killed. All spores killed. Do! Do. Do. AMERICAN FOULBROOD. 25 From Table III it will be noted that there were about 25,000 colonies in the check cultures, Wlien a similar suspension was heated only about 4,000 spores remained alive after 1 minute at 100° C.^ 148 after 5 minutes, 44 after 8 minutes, 14 after 10 minutes, and none after 11 minutes. The conclusion is reached, therefore, that the time element is a factor in the destruction of the spores of American foulbrood by heat. These results further show that the spores of different samples vary in their resistance to heat. It is interesting to note that by heating the disease material at 100° C. for one minute, more than 80 per cent of the spores were killed, and in 5 minutes, more than 99 per cent were destroyed. In Table IV are summarized results which indicate further the value of the time element in the destruction of the spores by heat. In this instance spore material from 6 different localities was used, the heat- ing being done at 95° C. with the spores suspended in water: Tabi,e IV. — Effect of the time element when 95° C. is iised in heating Amei-ican foulbrood spores Source of samples Temperature. Period of heating. Cuba. Colo. Ohio 4504. N. Y. Ohio 4519. Minn. "C. "F. Minutes. 95 203 12 _ _ _ _ _ _ 95 203 15 _ _ _ _ _ _ 95 203 20 — — — _ _ + 95 203 40 — — + + + + 95 203 50 — — + + + + 95 203 60 + + + + + + From Table IV it will be observed that the spores from all of the six samples studied resisted in an aqueous suspension a temperature of 95° C. for 15 minutes. Spores from the Minnesota sample were destroyed in 20 minutes at 95° C, those from the Ohio and the New York samples in 40 minutes ; and those from the Colorado and Cuban samples in 60 minutes. The time element is shown again to be an important factor in the destruction of American foulbrood spores. These results show also, as was shown above, that the spores con- tained in different samples vary as to their thermal death point. Other experiments were made showing the resistance of the spores of Bacillus larvae suspended in water to heat. At 93° C. (199° F.) spores from a Minnesota sample were destroyed in one hour, those from a Colorado sample in IJ hours, while those from the resistant Cuban sample were not all killed in 3 hours. Experiments using Cuban, Colorado, and Pennsylvania samples showed in each instance that numerous spores remained viable after 2 hours' heating at 90° C. (194° F.). 26 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. RESISTANCE OF AMERICAN FOULBROOD SPORES TO HEAT WHEN SUSPENDED IN HONEY The efifects of heating American foulbrood spores suspended in water are shown above. "When suspended in honey a different degree of resistance is to be expected. The technique used to obtain facts relative to such resistance is quite similar to that described (p. 22) for the experiments in which the spores were heated in aqueous sus- pensions; instead of water, however, the spores were suspended in honey ^ or honey diluted with water. For testing whether or not the spores had been killed by the heating, cultures were used in some instances and bees in others. In Table V are summarized some of the results obtained when the spores suspended in honey diluted with an equal volume of water were heated and tested by cultures: Table V. — American foulbrood spores heated in diluted honey ^ Temperature. Period of heating. "C. 'F. Minutes. 98 208 10 98 208 10 98 206 20 98 208 20 98 208 20 100 212 10 100 212 20 101 214 8 101 214 10 102 216 8 103 217 3 104 219 3 105 221 3 Origin of sample. Cultural results. April-May, 1915, Nebraska New York do Nebraska , Illinois New York do Washington State. Colorado Cuba Wasriington State, do Cuba Numerous spores not killed. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. About 500 spores not killed. > The altitude of the laboratory in Washington being nearly sea level the boiling temperature of water is almost 100° C. The higher temperatures recorded in this table and the following one were obtained by immersing the capsule containing the suspension in a solution of sodium chlorid and other salts. Table V shows that numerous spores were alive after being heated in a suspension of honey diluted with equal parts of water for 20 minutes at 100° C. The spores of American foulbrood are not destroyed by heat as readily, therefore, when suspended in diluted honey as when suspended in water. Experiments were made in which the spores were heated suspended in undiluted honey and tested by the cultural method. Table VI indicates the nature of the results obtained: ' The honey used was purchased at the market in sealed cans and was from the crop of the year preceding the experiments. J AMERICAN" FOULBKOOD. 27 Table VI. — American fonlhrood spores heated in imdilated honey Temperature. Period of heating. Origin of sample. Cultural results. April-May, 1915. "C. 100 "F. 212 212 212 212 212 212 217 221 221 225 225 225 Minutes. 10 10 10 20 20 20 25 20 20 20 30 40 Cuba Numerous spores not killed. Do. 100 Washington State 100 Ohio Do. 100 Cuba Do. 100 Washington State Do. 100 Ohio Do. 103 Washington State Do. 105 do Do. 105 Cuba Do. 107 do Do. 107 do Do. 107 do Do. From the results given in Table VI it will be observed that in every instance the spores suspended in honey resisted 100° C. for more than 10 minutes. It will be noted that numerous spores were alive after being so heated for 20 minutes. In fact they resisted 105° C. for 20 minutes and more. It is shown that numerous spores from the re- sistant Cuban sample were still alive after 40 minutes heating at 107° C. When suspended in honey, therefore, spores are much more re- sistant to heat than when susnended in water (p. 22) or when sus- pended in diluted honey (p. 26). Wliile the thermal death point has not been definitely determined for the spores suspended in honey, the results obtained indicate that the point was being approached by the experiments of Table VI. It was found that the spores suspended in honey were killed readily by heating in the autoclave at 15 pounds pressure, being destroyed, in fact, by the time this pressure was reached. In making these tests the suspension in test tubes was brought to 100° C. before being placed in the autoclave. After the heating cultures were made. SPORES HKATED IN HONKY AND FEB TO BEE.S A further set of experiments was made in which American f oulbrood material suspended in honey was heated to 100° C. and tested by the inoculation of colonies of bees. In performing the experiments a concentrated aqueous suspension of the disease material was added to hot honey, which is handled more readily than the cooler honey, until each 15 c.c. of honey suspension contained the material from 3 to 5 scales. This was then distributed in test tubes, each tube receiving about 15 c.c. of the suspension. These tubes were heated as were the aqueous suspensions in experiments reported above (p. 22) , by immersing them in water. After heating, the contents of the tube were added to sirup and colonies were inoculated. In Table VII some of the experiments performed are summarized : 28 BULLETIN 809, IT, S. DEPARTMENT OF AGRICULTURE. Table VII. — American foulhrood spores heated in honey and fed to bees Date of inoculation. Temperature. Period of heating. °C. °F. Minutes. 100 212 10 100 212 12 100 212 15 100 212 18 100 212 20 100 212 25 100 212 30 100 212 30 100 212 45 100 212 90 100 212 120 Results of inoculation. 1916 Aug. 25 Sept. 12 Do Do Aug. 25 Sept. 12 Aug. 11 Aug. 25 Do Oct. 11 Do American foulbrood produced. No disease produced. Do. Do. Do. Do. Do. Do. Do. Do. Do. It will be noted from Table VII that American foulbrood was produced after the scale material had been heated in honey for 10 minutes at 100° C, but was not produced when this temperature was maintained for 12 minutes or longer. The amount of heating required to prevent American foulbrood in the experiments recorded in this table is not as great as might have been expected from an examination of Table VI. The results tend to indicate that the vu'ulence might have been affected somewhat before the spores were killed. Sufficient data to prove the point, however, are yet lacking. From the various experiments relating to the thermal death point of American foulbrood spores recorded on the foregoing pages, it will be observed that 100° C. maintained for 10 minutes, with an exception now and then, is sufficient to kill all of the spores when sus- pended in water. It will be observed also that 98° C. applied for 10 minutes will do this in many instances and that 95° C. is sufficient in some cases. It is shown, furthermore, that a large majority of the spores are killed at 90° C. in 10 minutes and also that a very large number of them are killed at 100° C. in 1 minute. That a difference exists in the thermal death point of American foulbrood spores was expected. A difference is seen in the spores from different samples, in those from different larvaD or pupse in the same sample, and indeed in those of the same larvae or pupae. The maximum difference, when expressed in terms of degrees of tem- perature, between the most resistant spores of any two samples among those studied, the temperature being maintained for 10 minutes, is approximately 7° C. ; and when expressed in time, the temperature being maintained at 100° C, is approximately 7 minutes. These differences are for tho spores suspended in water. Suspended in honey a greater difference is to be expected. That still greater variations exist than those observed in these studies is most certain. To meet such differences the beekeeper has been urged to employ for the destruction of the spores in practical apiculture somewhat AMERICAN FOULBROOD. 29 more heat than the minimum amount that is required as determined by experimental studies. RESISTANCE OF AMERICAN FOULBROOD SPORES TO DRYING That the American foulbrood spores are able to withstand much drying has long been realized by beekeepers. Scales of American foulbrood obtained in 1907 from colonies in which the disease had been produced through experimental inoculation were stored in the laboratory. Each succeeding year for 9 years tests were made relative to the viability of the spores in this material. In 1916 they were still alive and as resistant to heat and as virulent as at any previous time. It is most likely that they would have with- stood the drying at room temperature for a very much longer period than the 9 years. RESISTANCE OF AMERICAN FOULBROOD SPORES TO DIRECT SUN- LIGHT RESISTANCE OF AMERICAN FOULBROOD SPORES WHEN DRY TO DIRECT SUNLIGHT In conducting experiments relative to the resistance possessed by spores of Bacillus larvae, when dry, to the direct rays of the sun, a heavy aqueous suspension is made of the scale material. This is spread in a thin film in Petri dishes and allowed to dry, the amount of disease material in each dish being equal to that in from 3 to 5 scales. The dry layer of disease material is then exposed to the direct rays of the sun and after different intervals of time cultures are made to determine whether the spores are viable. Table VIII gives a summary of the experiments made: Table VIII. — Resistance of spores of Bacillus larvae loTicn dry to the suri's rays Date of experiment. 1916 Aug. 31 Aug. 2S Sept.5 Do Do Sept. 6 Sept. 28 Sept. 19 Sept. 25 Aug. 23 Aug. 28 Sept. 23 Sept. 25 Sept. 16 Do Sept. 25 Period of exposure Hours. 2 4 5 7 10 11 12 29 38 28 37 41 41 44 61 79 Results as shown from cultures. Numerous spores not killed. Many spores not killed. Do. Several spores not killed. Few spores not killed. Do. Many spores not killed. Several spores not killed. A dozen spores not killed. All spores killed. Do. Do. Do. Do. Do. Do. From Table VIII it will be observed that aU of the spores were killed by the direct rays of the sun in from 2§ to 41 hours. It is natural to expect that the period required would depend upon the 30 BULLETIN" 809, V. S. DEPARTMENT OF AGRICULTURE. date and time of exposure and somewhat upon the thickness of the film of disease material exposed. The climatic conditions during August and September, the time of the experiments, were naturally favorable for the destruction of bacteria by sunlight. The expos- ures were made only when the sky was clear, preference being given to the middle portion of the day. REISISTANCE OF AMERICAN FOULBROOD SPORES WHEN SUSPENDED IN HONEY TO DIRECT SUNLIGHT The technique used is as follows : A concentrated aqueous suspen- sion of spore-containing material is added to honey in Petri dishes, each dish receiving the disease material equal to that of from 3 to 5 scales. These are exposed to the direct rays of the sun. The tops are used with the dishes to prevent robbing by bees. After different periods of time the contents of a single dish are added to sirup and fed to a colony free from disease. A set of experiments made is summarized in Table IX. Table IX. — Spores of Amrricnn foidhrood in honey exposed to the sun Date of exper- imeut. Period of expo- sure. Results of inoculation. 1916. July 13 July 15 Hours. 6 13 Large amount of American foulbrood produced. Moderate amount ol disease produced. July 26 Aug. 25 Sept. 1 Aug. 10 Aug. IS Sept. 14 Do. Tl'fffcs. 2 4 5 4 5 6 8 Scattering cells of diseased brood. Considerable diseased brood. Do. No disease produced. Do. Do. Do. The data given in Table IX show that the spores of American foulbrood w^hen suspended in honey were destroyed by the direct rays of the sun in from 4 to 6 weeks. As the suspension in each dish contained the disease material of from 3 to 5 scales it is evident from the results obtained that many of the spores must have been destroyed in a comparatively short period. It will be readily appreciated that here again the period required for the destruction of the spores will vary greatly with the intensity of the sun's rays to which the honey suspension is subjected. In these preliminary experiments the period of exposure represents the entire time from the beginning of the exposure to the time of inocu- lation. There were days during the exposure on which the sun shone a great deal, others on which it shone very little, and stiU others on which it did not shine at aU. The comparatively low tem- perature of the honey suspension attained during the exposure to the sun could not have been an important factor in the destruction of the spores. AMERICAN" FOULBROOD. 31 RESISTANCE OF AMERICAN FOULBROOD SPORES TO FERMENTATION A few preliminary experiments relative to the effect of fermenta- tive processes on the spores of American foulbrood were made. The processes involved concern chiefly the sugar splitting and the proteo- Ijrtic enzymes. In experiments relative to the former, the scale material was suspended in a 20 per cent aqueous solution of honey and, in those relative to the latter, it was suspended in a 1 per cent aqueous solution of peptone. In each case a small bit of soU was added to inoculate the suspending solution still further. These suspensions, respectively, were distributed in test tubes and allowed to undergo fermentation. Observations were made at incubator and outdoor temperatiu-es. The outdoor temperature was that which obtained in an empty hive body covered, and standing in the apiary at the time of the year shown by the dates of the experiments. After intervals reckoned in days colonies free from disease were inoculated, each with a suspension from a single tube. The experiments are summarized in Table X: Table X. — Resistance of the spores of Bacillus larvae to feriacntation Date of experi- ment. Nature of the suspension. Period. Temperature. Results of inoculations. 1916. Aug. 14.. Honey Days. 26 53 24 53 26 53 24 53 Outdoors American foulbrood produced. Do. Sept. 9... Aug. 11.. Sept. 9.. Aug. 14.. Sept. 9.. Aug. 11. do do do Incubator Do. do do Do. Outdoors Do. do . . do. .. Do do do Incubator Do- Sept. ... do Do. From Table X it will be seen that in the presence of fermentative processes in the honey solution and in the peptone solution the spores were alive and virulent after 53 days at incubator and at outdoor temperatures. It is quite probable that they would have remained so for a very much longer period. RESISTANCE OF AMERICAN FOULBROOD SPORES TO CHEMICAL DISINFECTANTS While a mass of data was obtained relative to the effect of chemical agents on the spores of American foulbrood, the results are still considered as being preliminary in nature. In conducting the experiments a suspension of disease material in an aqueous solution of the disinfectant was used. This was drawn into capsules ^ (fig. 9) > The ampules used are made from glass tubing of small bore by heating and drawing it into three bulbs. The two intermediate constricted portions are wrapped with cotton. After being sterilized the ampules are filled with the spore-containing suspension sealed and placed at different temperatures. Incubator, room, and outdoor temperatures were employed. At the time of making the cultures the ampule is broken at the two intermediate constrictions and the spore-containing suspension used in making the inocu- lations is obtained from the intermediate bulb. By this method an immersion of the spores in the solution for the entire period is assured. 32 BULLETIN 809, U. S. DEPARTMENT OF AGRICULTURE. and sealed. After different intervals of time the suspension was cultured, using brood-filtrate-egg-yolk-suspension agar. The results show that spores of American foulbrood were alive after being in suspension in 1, 2^, and 5 per cent aqueous solutions of carbolic acid (commercial) for months at room, outdoor, and ice- box temperatures, respectively. The maximum period during which any suspension was kept at ice-box temperature was 32 months. At incubator temper- ature the period during which the spores remain alive in 5 per cent carbolic acid is better reckoned in weeks than months. In 5 and 10 per cent solutions, re- spectively, of formalin (37 per cent formaldehyde) it was found that they were viable after 6 hours, and in a 20 per cent solution they were alive after 30 minutes. Wliile the spores were alive in the formalin suspensions after these periods, the evidences obtained indicate that the chemical death point was being reached. Mercuric chlorid was resisted for days in 1 to 1000 and 1 to 500 solutions, respectively. Inasmuch as the spores of American foulbrood resist the action of carbolic acid for more than a month under fairly favorable conditions for then- destruction, this agent could scarcely be used with profit in prac- tical apiculture. As the destruction of the spores with formalin is a question of hours only, it is possible in well-selected cases that use could be made of this agent. Since mercuric chlorid in strengths ordinarily used is resisted by the spores for days and furthermore is such a violent poison, its value as a disinfectant in American foulbrood is evidently limited. The general Fig 9 — car)suie used conclusion In regard to chemicals as disinfectants is, in determining the therefore, that they offer very little promise as a resistance of spores j.-i £ J.^ j j. j.- rj.i e to chemical disin- practical means tor the destruction oi the spores oi fectants. American foulbrood. EFFECT OF DRUGS ON AMERICAN FOULBROOD Divers experiences have been reported by beekeepers relative to the value of drugs in the treatment of the bee diseases. The fact that the spores of Bacillus larvae possess a marked resistance to chem- ical disinfectants tends to discourage hope that such substances would be of much therapeutic value when employed as drugs in American foulbrood. Such a conclusion, naturally, would not foUow necessarily. Some data relative to the effects of drugs on this disease have been obtained through experimental inoculations using beta- naphthol, U. S. P.; carbolic acid (phenol), C. P.; oil of eucalyptus, AMERICAN FOULBROOD. 33 U. S. p.; formic acid, C. P.; salicylic acid, U. S. P.; salol (phenyl salicylate, U. S. P.); and quinin (bisulphate of quinin, U. S. P.). In making the inoculations a suspension of scale material in water is added to medicated honey and colonies free from the disease are fed. The different drugs are used in different proportions. Honey, rather than sirup, is employed since the bees will take the drugs in higher proportions when in honey than when in sirup. The strength that they will take varies somewhat with the conditions present. The higher proportions recorded in Table XI summarizing the experiments approximate the maximum amount that can be em- ployed : Table XI. — hidicating the effect of drugs on American foulh rood Date of experi- ment. Drugs. Strength. Results of inoculation. 1916. July 11 May 29 Betanaphthol 1:2000 1:1000 2:1000 1:2000 1:1000 2:1000 10:1000 4:1000 3:1000 1:2000 1:1000 2:1000 1:2000 1:1000 2:1000 2:1000 3:1000 10:1000 1 do do July 11 June 20 Carbolic acid do Do.. do May 29 Do.. do Oil of eucalyptus July 11 Do.. Formalin [American foulbrood produced. May 29 do June 5 do July 11 Salol May 29 do June 5 do July 11 May 29 do June 5 do From Table XI it will be seen that American foulbrood was pro- duced in all cases in which colonies were fed a suspension of diseased material in honey medicated with betanaphthol, carbolic acid, eucalyptus, formic acid, salicylic acid, salol, and quinin respectively in the proportions noted. In some of the experiments medicated sirup free from the spores was fed to the inoculated colony on a few successive days following the inoculation, and in some instances both preceding and following it. Whether these treatments with the medicated sirup produced any effect on the infection — positive or negative — was not determined definitely. The results thus far obtained indicate that the drugs cited here can not be depended upon, for the present at least, in the treatment of American foulbrood. They do not preclude the possibility, however, that other drugs might be used with profit, but they do emphasize the fact that beekeepers should make sure that the value of a drug has been clearly demonstrated before it is used. 132862°— 19— Bull. 809 3 34 BULLETIN 809^ U. S. DEPARTMENT OF AGRICULTURE. The results recorded on the foregomg pages relative to the resis- tance of American foulbrood spores to heat, drying, fermentation, sunlight, chemical disinfectants, and drugs, it will be observed, are only approximate from a strictly technical viewpoint. For practical purposes, however, they are in most instances entirely adequate. In using any of these results in deVising means for the destruction of the spores in practical apiculture the time element determined by the experiments should be somewhat mcreased in each instance. MODES OP TRANSMISSION Observations made during these studies point to certain paths by which American foulbrood is most likely to be transmitted. The evidences obtained tend to support certain views which have been entertained by beekeepers and to negative others which have been suspected by some as possible. Inasmuch as the disease may be produced in larvae by feeding a colony Bacillus larvae in pure cultures, or from decaying remains of brood dead of the disease, it would seem that the portal of entry of the virus is somewhere along the alimentary tract of the larva. The fact leads at once to the sus- picion that the food of bees contaminated with disease material is a very probable source of infection. Were the water supply likewise contaminated naturally it also would be a probable source of infection. Two prerequisites for the appearance of American foul- brood in a colony are (a) larvae of the feeding age, and (6) a suffi- cient amount of disease material in the food or water supply of bees. A small amount of brood or a heavy flow of nectar, therefore, would tend to reduce the likelihood of colony infection under con- ditions otherwise favorable for infection. Bees manifest a tendency to remove brood dead of American foulbrood as shown by the remains of partially removed larvae and pupae dead of the disease. The removal is done piecemeal and is accomplished more readily when the brood is recently dead than when the decaying remains are at all viscid or dry. Were the fate of the removed fragments definitely known much more could be said concerning the spread of the disease in the colony. It has been observed that a small amount of infection — a dead larva or pupa here and there — may be present in a colony for months, a year, and even longer, without causing a heavy infection in it. There is considerable evidence to support the belief that occasionally in cases of light infection the disease may disappear unaided by treat- ment. Such a phenomenon frequently takes place in sacbrood and indeed should be expected to occur now and then in American foulbrood. It should be emphasized that such a course for the disease, if it occurs at all, is unusual. Although American foulbrood mmmm UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 810 Contribntion from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER February 26, 1920 EUROPEAN FOULBROOD By G. F. WHITE, Specialist in Insect Diseases CONTENTS Page Introduction 1 Name of the Disease 2 Healthy Larvae of the Age at which they Die of European Fonlbrood .... 3 Symptoms 4 Etiology 7 Technique 15 Thermal Death Point of Bacillus pluton 17 Resistance of Bacillus pluton to Drying . 17 Resistance of Bacillus pluton when Dry to Direct Sunlight 19 Resistance of Bacillus pluton in Water to Direct Sunlight 20 Resistance of Bacillus pluton in Honey to Direct Sunlight 21 Pago Resistance of Bacillus pluton to Fermen- tation Resistance of Bacillus pluton to Putrefac- tion Viability of Bacillus pluton in Honey . . Viability of Bacillus pluton in Pollen . . Resistance of Bacillus pluton to Carbolic Acid Eifect of Drugs on European Foulbrood Transmission of European Foulbrood . Diagnosis 28 Prognosis 31 Summary and Conclusions 31 Literature Cited 34 Explanation of Plates 37 21 22 23 24 24 26 26 WASHINGTON GOVERNMENT PRINTINa OFFICE 1920 UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 810 Contribution from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER February 26, 1920 EUROPEAN FOULBROOD By G. F. White Specialist in Insect Diseases Introduction Name of the disease Healthy larvas of the age at which they die of European foulbrood — Symptoms Etiology Technique Thermal death point of Bacillus pluton Resistance of Bacillus pluton to drying Resistance of Bacillus pluton when dry to direct sunlight Resistance of Bacillus pluton in water to direct sunlight Resistance of Bacillus pluton in honey to direct sunlight CONTENTS Page 2 3 4 7 15 17 17 19 20 21 Resistance of Bacillus pluton to fer- menta.tion Resistance of Bacillus pluton to putrefaction Viability ot Bacillus pluton in honey_. Viability of Bacillus pluton in pollen- Resistance of Bacillus pluton to car- bolic acid Effect of drugs on European foul- brood Transmission of European foulbrood- Diagnosis Prognosis Summary and conclusions Literature cited Explanation of plates — Page 21 22 23 24 26 26 28 31 31 34 37 INTRODUCTION European foulbrood is an infectious disease of the brood of bees caused by Bacillus pluton. It is characterized by the death of brood during its uncapped stage and by the absence of any marked odor. The disorder has a wide distribution and is fairly well known to bee- keepers. The losses sustained by the infected apiary vary from a slight weakening of the colonies in some instances to the destruction of all of them in others. Practical apiarists have determined mucn concerning the disorder while pursuing their profession. The writer in an earlier paper (15)^ referred to the nature and extent of the progress that had been made in the study of the disease from the laboratory point of view. The present paper deals with results which have been obtained from a con- tinuation of the work. Among the problems considered are : The re- ^ Figures in parentheses refer to " Literature cited," p. 34. 132817'— 20— Bull. 810 1 2 BrLLETIN 810, U. S. DEPARTMENT OF AGRICULTUEE. sistance of Bacillus pluton to heat, diving, sunlight, fermentation, and disinfectants; the effect of the disease on the cokjny and on the apiary ; and the transmission, diagnosis, and prognosis of the disease. Work directly on the treatment of the disease has not been attempted by the writer. N'aturally, however, any treatment that is devised, if it is to be efficient and at the same time economical, must be based upon results obtained from the solution of such problems as those which have received attention in these studies. Eesults obtained from a study of the disease in the laboratory and in the experimental apiary form the basis of the discussions contained in the present paper. Since the disease encountered in nature is very similar to the one produced by artificial inoculation, the importance of the studies is at once evident. The paper ^ will be of interest, it is believed, not only to the apiarist who may wish to apply the facts here determined in the pursuit of his profession, but also to the investigator whose desire primarily is a further study of the disease. NAME OF THE DISEASE The term " foulbrood " was quite generally used in the past, as it still frequently is, for the two infectious diseases now known in America as European foulbrood and American foulbrood. In 1885 when Cheshire and Cheyne (4) in England made their studies on foulbrood and described Bacillus alvei, evidently they w^ere not con- vinced that there were two distinct diseases that were being called by the one name foulbrood. The disease studied by them is the one which is the subject of discussion in the present paper. In the names for the two diseases it Avill be observed that the word " foulbrood " is retained in both instances. To this " European " is added for the disease on which early laboratory studies were made by these Euro- peans (Cheshire and Cheyne). Dr. William R. Howard (6), of Texas, in 1900, worked for a brief period Avith this disease, reached the conclusion that it was a new one, and referred to it by the names " New York bee disease," or " black brood." Work by Moore and White (11) in 1902 showed that the disease was not new, but Avas the foulbrood studied by Cheshire and Cheyne (4). The names " New York bee disease," or " black brood," therefore, were superfluous, and as their use would have added to the confusion that already existed they were discarded. Beekeepers, ento- mologists, and pathologists, as a rule, are more or less familiar with the terms " foulbrood " and " Bacillus alveiy Usually, however, the ropy foulbrood— American foulbrood— is the one that is thought of, 'The prosent studies arc similar to those made by the writer on sacbrood (17), Nosema- disease (18), and American foulbrood (19). A reference to these papers may be found helpful whore the discussions in the present one are especially brief. The investigations were completed in September, 1016, jind the paper was submitted for publication in October, liiis. EUROPEAN FOULBROOD. 6 and the one that frequently has been associated in the literature witli Bacillus alvei This is unfortunate. While B. alvei is not the cause of any bee disease, it occurs very frequently with European foul- brood and is found only seldom in the ropy disease. In using the names European foulbrood and American foulbrood it is possible, however, to avoid confusion by bearing well in mind the history of the disease. HEALTHY LARV^ OF THE AGE AT WHICH THEY DIE OF EUROPEAN FOULBROOD Bees dying of European foulbrood do so during the larval stage.^ Death may take place nt any time from the fourth day of larval life to pupation. For convenience of description the brood of the age at which death from European foulbrood occurs is placed here in three groups. Groups 1 and 2 include the uncapped and group 3 the capped larvae. GROUP 1 The youngest larva (PI. II, D, G) that dies of European foulbrood practically covers the bottom of the cell. It lies either on its right or its left side, with its dorsal portion extending to the lateral walls of the cell. Its form is C shaped with the anterior and posterior ex- tremities almost together. Its color is bluish white with a glistening surface, presenting a pearly appearance. The body is more or less opaque, due largely to the adipose tissue. Folds and furrows divide the surface into segments. In health these are quite prominent and the entire larva is turgid in appearance. With the unaided eye spiracles and tracheae can be seen with diffi- culty, but by slight magnification they are readily observed. Most of the tracheae, appearing as white lines, extend either dorsally or ventrally on the lateral side of the larva, but a distinct chain con- necting them will be observed to extend at right angles to these. GROUP 2 Healthy larvae (PI. Ill, D, G) slightly older than those described in Group 1 constitute Group 2. The larva now completely fills the bottom of the cell. The dorsal side pressing against the lateral side walls of the cell causes the contour of the body to be in general hexagonal. The tracheae are seen less easily than in younger larvae, while the color, glistening appearance, prominence of segments, and turgidity are similar to those of the younger larvae described in Group 1. By turning the larva so that its dorsal surface may be brought into view (PL III, A) there is observed a more or less transparent narrow I ' The term larvK as used in the present paper applies to the prepupae as well as to earlior stages of the brood. 4 BULLETIN 810, U. S. DEPAETMEXT OF AGRICULTURE. area along the dorsal median line extending nearly the length of the body. The contents of the stomach may be seen through this area. The color of the mass is due chiefly to the presence of pollen. It is usually some shade of yellow. The median area presents in its appearance a sharp contrast to the bluish-white, opaque portions on either side of it. Similar appearances are to be noted in the larvae of Group 1. The larva removed from the cell performs only slight movements, lies partly coiled, and is more or less turgid. The segments are promi- nent. When the body wall is torn there flows from the ruptured wall the clear larval blood, in which are suspended often fat and other tissue cells which give to it a somewhat milky appearance. The stomach, a transparent tube easily torn into segments, contains a mass of partially digested food, pollen constituting usually a con- spicuous portion of it. GROUP 3 Group 3 consists of capped larvae. These are, therefore, larger than those described in Groups 1 and 2. In the group are included the larvae which have spun a cocoon as well as those which have not. An endwise position in the cell may or may not have been assumed. The larvae are seen in various positions. Not infrequently some portion of the dorsal surface is turned toward the observer, the narrow, me- dian, transparent area being in evidence as in younger larvae. Healthy larvae occupying an endwise position are described in papers on sac- brood and American foulbrood (17, 19) and Avill not be referred to further at this time. SYMPTOMS In European foulbrood, as in other brood diseases, the colony as a whole and not the individual bee should be considered as the unit in the discussion of the symptoms of the disease. The description of the symptoms recorded in the present paper is based chiefly upon observa- tions made on the disease produced through artificial inoculations. In making the studies in the experimental apiary observations made by beekeepers have been duplicated and new facts determined. It has been possible also to locate errors which have been made in discussions of symptoms of the disease. GENERAL SYMPTOMS FROM A CASUAL EXAMINATION Death of brood during the feeding stage, in uncapped cells, is a characteristic of European foulbrood. The brood nest in the disease usually presents an irregular appearance, capped cells and uncapped ones being found scattered irregularly over the brood frames, giving to them the " pepper box " appeai-ance (PI. I) often referred to by EUROPEAN FOULBROOD. 5 beekeepers, a condition noticeable when the disease is fairly well advanced in the colony. The dead larva> lose their pearly whiteness and assume a yellowish color, later becoming brownish. This deepens often to a dark brown. The decaying remains are not characteristically ropy, as in American foiilbrood. Marked viscidity is usually absent. When it is present the decaying mass can be drawn into threads but to a less extent than in the ropy disease. In advanced cases the disease may be accom- panied by an odor, but in the writer's experience this never has been marked and never offensive. As the disease in the colony advances, weakness becomes a symp- tom. In severe cases queenlessness may result from the infection. This, however, is by no means the rule. SYMPTOMS MANIFESTED BY INDIVIDUAL LARV^ SICK OR DEAD OF EUROPEAN FOULBROOD Evidences of European foulbrood in the individual larvae appear before and after death. The colony symptoms used most frequently in the diagnosis of the disease are largely post-mortem appearances of larvae. Of much interest and frequently of considerable diagnos- tic value are the symptoms manifest by larvae sick but not dead of the disease. For convenience in the description of the appearances of the sick or dead larvae, the grouping used in describing the healthy larvae (p. 3) is followed. The appearances of affected larvae both living and dead are, of course, changing constantly. A description which is correct for one day or hour, it should be realized, is not likely to be entirely correct for the next. The youngest larvae manifesting symptoms of European foul- brood are approximately 4 days old (PI. II, A, B, C, E, F, H, I). In many cases at this stage of the disease a peristalsis-like movement of the body is marked and is readily observed by the unaided eye, but in others no such bodily movements are observed. The diseased larvae at the time may be more transparent (PI. II, B, H) than healthy ones of the same size. In such larvae the tracheae are quite prominent and more readily seen than in healthy ones. Occasionally numerous minute opaque areas are observed in these more transpar- ent larvae, giving to them a punctate appearance. Very often, how- ever, this sign is not present. In many instances, indeed, no distinct symptom is observed until the larva approaches death. (PI. II, A). Larvae (PL II, A, B, C, E, H, I) of this group dying or just dead of the disease lose their marked glistening appearance ; their pearly whiteness gives way to a yellowish tint ; the turgidity seen in healthy larvae is diminished in the sick; and the folds and furrows indicat- G BITLLETIX 810, IT. S. DEPARTMENT OF AGRICULTURE. iiig the segments of the body become less prominent. As the process of deca}' advances the yellowish hue changes, the color assuming a brownish tone. The segmental markings are less prominent, while the tracheae often become quite distinct, appearing as white lines contrasted with the darker color of the larval remains (PI. II, B). Xot infrequently at this time there will be seen a chitinous envelope containing a watery-looking fluid in which is the larva proper (PI. II, C ; PI. IV, A). The decay proceeds and the drying becomes evi- dent. The larval mass settling upon the concave bottom of the cell causes the upper surface of the mass to be depressed about the cen- ter. At this stage the tracheae not infrequently are seen distinctly in the drying mass. When the larval remains become dry they are known as the scale (PI. II, F). The scales do not adhere closely to the cell and when removed are found to be thin and more or less circular in outline. They are convex and smooth on the side which was in contact with tlie bottom of the cell while the opposite surface — the one which, while in the cell, was toward the observer — is slighth'- roughened and concave. Larva3 (PI. Ill) showing symjDtoms of European foulbrood and classed in this group have reached a sufficient size to fill the deepest third or more of the cell. The yellowish tint appeai-s in contrast to the bluish white of the healthy lan-se (PI. Ill, D, G) . Increased movement may or may not be observed. Before and after death the remains may assume one of a number of positions in the cell. Not infre- quently a portion of the dorsal surface is turned toward the observer (PI. Ill, B). Usually through the transparent area along the me- dian dorsal line a whitish or yellowish- white mass is to be observed. This mass is within the stomach of the larva and contains a large amount of bacterial growth (PI. VIII, a, b, c) consisting very largely of Bacillus pluton. Often before death this mass is seen to move within the stomach in response to the peristalsis-like movements of the body of the larva. At the time of death the larva usually occupies some unnatural position, being more or less curled up and lying upon the floor of the cell (PI. Ill, C, E, F, H, I). Lessened turgidity, a relative dullness of the surface appearance, and a yellowish tint are present, l^ot in- frequently the two ends of the larva are directed more or less toward the bottom of the cell and some portion of the dorsal surface is toward the opening of it (PL III, E, H, I). Among the dead larva; will be found some with one end directed toward the bot- tom, and the other toward the mouth of the cell, the body occupying a more or less spiral position against the side walls and floor of the cell (PI. Ill, F). EUROPEAN FOULBROOD. 7 Later the dead larval remains assume a brownish tint which deepens to varying shades as decay continues and drying takes place. During the early part of the decay, the firmness of the body wall per- mits the removal of the larva intact from the cell. Later, however, it offers but little resistance and is easily ruptured. The decaying mass before drying often attains a certain amount of viscidity. Sometimes it is of a doughy consistency, at other times it is purulent or sputum- like, while at times it assumes a viscidity that will permit of its being drawn out to the extent of an inch or more. When the larval mass becomes dry it forms an irregular scale, usually brown in color, lying on the floor or side wall of the cell or both, but not adhering closely to them. GROUP 3 A larva dying of European foulbrood after benig capped may be found occupying one of many positions within the cell (PI. IV, C, D, E ; PI. V, D, E, F, G, H). Dying before the two-day quiescent period that precedes pupation, the remains during decay and as a scale resemble in many respects those of larva? described in group 2. The dry scales occupy usually an irregular position on the floor of the cell (PI. IV, F, G). Dying during the two-day quiescent period, however, the scales (PI. V, F, I) resemble very much those of larvae dying at the same age of American foulbrood. The larval mass assumes the brownish hue which deepens as the decay advances, reaching a dark brown. Viscidity is present in the decaying larval mass, but the extent to which the decaying material may be drawn out is less than in American foulbrood. The scale is less brittle and more rubberlike. At no time has the writer observed pupse dead of European foul- brood. If they die of the disease it is a rare occurrence. The removal of larvae sick or dead of the disease is accomplished to a greater or less degree by adult workers. The larvte are either partially or entirely removed. This is usually done x^iecemeal. In an infected colony will be found, therefore, the remains of larvae of different ages (PI. IV, B) and (PI. V, A) in varying numbers. ETIOLOGY PREDISPOSING CAUSES Age. — Infection in European foulbrood takes place during the feeding stage and at some time after the first day of larval life, the larvae being more often 2 days of age, or older. Death takes place somewhat more than 2 days from the time of infection. As a rule, therefore, a larva has passed its fourth day of larval life before death from European foulbrood occurs. From this age to pupation larvae may die of the disease. The writer has not encoun- 8 BL^LLETIN 810, U. S. DEPARTMENT OF AGRICULTURE. tered death among brood which has reached the pupal stage. Adult bees are not susceptible to infection. Sex. — "Worker, drone, and queen larvae are all susceptible to in- fection with Bacillus pluton and any of these may die of European foulbrood. l^ace. — Complete immunity from European foulbrood has not been found among the races of bees studied. Experimental work recorded in the present paper involved the use of at least five colonies of " tested Italians," two of " tested Carniolans," and two of " tested Caucasians." For the most part the bees used were " un- tested Italians," but among the colonies were a few common blacks. In all these strains the disease was readily produced through experi- mental inoculation. The examination of numerous samples of dis- eased brood received from beekeepers throughout the United States suggests that all races connnonly kept by American beekepeers are susceptible to European foulbrood. The relative immunity of the different races has not been demonstrated by the studies. These facts, however, do not dispute the observation by practical bee- keepers that some strains of bees show a greater colony resistance than others. Climate. — From reports of studies made in Austria by Muck (12), in Denmark by Bahr (1), in England by Cheshire and Cheyne (4), in Germany by Zander (20), and in Switzerland by Burri (3), it is clearly evident that the disease discussed in the present paper occurs in these different countries. It has been encountered also in many sections of the United States and Canada. This distribution shows that the infection can exist under a variety of climatic conditions. The practical import of the fact is that the presence of European foulbrood in any locality can not be attributed entirely to the climate of the region. Season. — Beekeepers have observed that European foulbrood oc- curs with greatest severity before midsummer rather than later in the season. The disease, it has been shown experimentally, can be produced, however, at any season of the year at which brood is being reared. Its severity at any given season is to be attributed, there- fore, to environmental conditions rather than to the difference in the susceptibility of larvae during the different seasons. Food. — As in American foulbrood it is found that the cause of the disease in the colony is governed very little if at all by the quality of food gathered by bees. Indirectly, however, the quantity present in the hive or obtainable often does influence its course materially. EXCITING CAUSE That Bacillus alvei may be present in large numbers in brood dead of foulbrood was demonstrated by Cheshire and Cheyne (4) in 1885. EUROPEAN FOULBROOD. 9 For a decade and a half following the observation the belief was quite general that this bacterium was the exciting cause of a bee dis- ease. The view was then seriously challenged. In l&OG the only positive conclusion in regard to the relation between European foul- brood and Bacillus alvei that could be draw^n by the writer (13) was that this species occurs in brood dead of the disease. William E. How^ard (6), of Texas, after a brief study of the dis- ease reported in 1900 the presence of an organism which he called Bacillufi millii. He cultivated the species apparently with ease. In 1904 Bahr (1) in Denmark found a small oval bacterium in a brood disease in which larvae clying in uncapped cells are yellowish in color and not ropy in consistency. Burri (3) in 1906 encountered in his studies on the brood diseases a small bacterium which he referred to as guntkeri-iorms. The species was cultured and compared with BacteHum guntheri and found to be somewhat different. In 1907 Maassen (7) obtained from brood material cultures of a species which he named Streptococcus apis. White (14) in 1908 reported the presence of a small organism in European foulbrood which had refused to grow on artificial media. The species was not the one, therefore, with which the investigators just referred to had worked. That this organism might be the exciting cause of the disease was noted. Pending more information regarding it, the species was not given a name but was referred to as bacillus " Y." That this species bears a direct etiological relation to the disease was demonstrated in 1912 by the writer (15) and the name Baoillus plnton was then given to it. As the cultivation of Bacillus pluton on artificial media had not been accomplished the conclusion that it is the exciting cause of European foulbrood was arrived at by eliminating all other possible agencies. The observations furnishing the proof appear in an earlier paper (15). By demonstrating Bacillius pluton to be the cause of the disease, Bacillus alvei., Streptococcus apis., Bacterium, eurydice., and Bacillus orpheus., and still other species occasionally encountered, were thereby proven to be secondary invaders. To eliminate the possibility of a filterable virus in European foul- brood 10 colonies were inoculated with filtrates obtained from aque- ous suspensions of brood sick and dead of the disease. In six instances the Berkefeld N filter was used and in four the Pasteur- Chamberland F was employed. In no case was the disease produced. Studies recorded in the present paper on the resistance of Bacillus pluton to heating, drying, fermentation, and disinfectants show that when the virus of the disease is not destroyed this species is still alive. This fact is further evidence in support of the conclusion that the species Bacillus pluton is the virus of the disease. 132817°— 20— Bull. 810 2 10 BULLETIN 810, U. S. DEPAKTMENT OF AGRICULTURE. BACILLUS PLUTON An artificial medium for the cultivation of Bwcillm iDluton has not vet been devised. To accomplish this may or may not be a particu- larly difficult task. The media ordinarily used in the laboratory are not suitable. Bee-larva? agar, brood-filtrate media, egg-yolk-sus- pension agar (19), and combinations of these have not thus far proved sufficient for the purpose. Tlie species is an unusual one. The generic classification has not been determined definitely and this may not be possible until the proper condition for the artificial culti- vation of the species has been supplied. The morphology of Bacillus -pluton is somewhat variable. In very early infection its form is that of a short rod in pairs or in chains, or possibly of a coccus with the individuals similarly arranged (fig. 1; PI. Vli, B). The length is then equal to or somewhat greater than the breadth. In slightly later stages of infection the predomi- nating form is that of a lancet- shaped coccus (fig. 1; PI. VII, A), and in late stages this form is present almost exclusively. The lancet form occurs singly, varj'ing greatly in size and hav- ing a length which approximates twice the width. The length is more often less than 1 pi than greater. The organism colors uniformly with the aniline stains, FIG. Z^^^BMs pluton. ^taius with irou hematoxylin, and is gram-positive. It does not form spores. This is evidenced by the microscopic appearance and also by the thermal death point of the species. Its resistance to dry- ing, disinfectants, and other environments is discu^ssed later in the present paper. Seven rabbits inoculated, six subcutaneous] y and one intraperi- toneally, with a suspension of larvae dead of European foulbrood proved to be refractory. Only a slight rise of temperature followed the inoculations and the weight was not materially affected. Six guinea pigs inoculated subcutaneously with similar material proved not to be susceptible to infection with the species. Four pigeons inoculated in the pectoral muscles and two white rats inoculated sub- cutaneously also proved refractory. In none of these inoculated ani- mals were there any lesions of particular note produced. Orowth of Bacillufi pluton in the infected larva begins close to the surface of the peritrophic membrane (PL VTI, I) in contact with the food of the larva. As growth continues the bacterial mass extends toward the center of the lumen of the peritrophic sac (PL VII, K), EUROPEAN FOULBROOD. 11 finally filling it more or less completely ( PI. VII, J ) . The growth does not always take place unifonnly along the peritrophic membrane (PL VII, J) , nor does it extend beyond it (PI. VII, I, J, K) , but is inclosed within the sac, the tissues of the larvae not being reached. The mul- tiplication of the organism after the death of the host, if, indeed, it takes place at all, is limited. Secondary invaders, chiefly Bacillus alvel^ BacteHwm eurydice^ Streptococcus apis^ and occasionally Bacillus orpheus^ and a few others, are encountered at various stages of the disease and during the decay of the larva. During the life of the larva these species also remain within the peritrophic sac. BACILLUS ALVEI Fig. 2, Bacillus alvei (fig. 2; PI. VII, D, F) is present very frequently and in very large numbers in larvse dead of European foulbrood. The species was well described by Cheyne (4). Descriptions maybe found elsewhere also (11, 13). It is readily recognized and may be differentiated easily from other spore - producing species occa- sionally encountered in the dis- eased brood. Bacillus alvei is not the active cause of any bee disease. It seems probable, however, that it plays a role in European foul- brood, but the extent is not fully known. The species is present usually, if not invariably, in large numbers in the rubberlike scales (PI. V, F, I), which resemble so much those of American foulbrood. The decayed larval mass, which forms the scale, before becoming dry is ropy in consistency similar to that of American foulbrood but to a less degree. It seems probable that this ropiness is due more or less directly to Bacillus alvei On account of this viscidity the decaying mass, as well as the scales, are removed with greater difficulty than are most larvse dead of European foulbrood. The result, as often observed, is that these brown viscid decaying larvae or the rubberlike scales result- ing from them are the only evidence that European foulbrood is present in the colony. While Bacillus pluton in such larval masses and scales is often diffi- cult to detect microscopically, its presence can be demonstrated through the experimental inoculation of healthy larva?. Inasmuch as Bacillus pluton will live for a considerable period in the scales, it liaxHllus alvei. Spores free from and others within rods. 12 BULLETIN 810, U. S. DEPARTMENT OF AGRICULTURE. seems quite probable that the disease might in some cases be carried over for months or even over winter through the medium of these rubberlike scales. It is of interest to know that the amount of disease resulting imme- diately from inoculations in which scale material is used is much less than when larvae recently dead of the disease are used. This is true also of dead larvse stored in Petri dishes compared with smears allowed to dry immediately from larvae recently dead of the disease. These facts indicate a possible deleterious effect on Bacillus pluton of the sec- ondary invaders multiplying in the decaying larvae. STKEPTOCOCCUvS APIS It is most probable that Streptococcus apis is the species that was isolated from diseased brood by Burri (3) and referred to by him in 1906 as " giintheri-f orms.''^ Maas- sen described it in 1908 (8). The organism grows well at incubator, room, and refrigerator tempera- tures in most of the media ordi- narily used in the laboratory. Its cultural characteristics suggest the micrococci rather than the streptococci. Confusion in some of the earlier investigations was due evidently to the resemblance of Streptococcus apis and Bacillus pluton morphologically. To this FIG. s.-streptococcus apis. fact is duc the chief interest in the species Streptococcus apis. Wlien encountered in larvae dead of European foulbrood it can be identified readily by culturing. The generic position of this species should be considered as being not altogether certain. Occtirrence. — Streptococcus api^ is occasionally encountered in larvse dead oE European foulbrood and often is present in larse numbers. Morpholoyy. — It is more or less spherical (fig. 3; PI. VII, E), occurring singly and in pairs with occasionally a chain of 2 or more pairs when grown in liquid media. In larval remains not infrequently the ends may be somewhat pointed. Staining properties. — It colors uniformly and readily with the common stains, and retains the stain after Gram's method. Obieose agar p7ofc.— Within a day growth is visible. Colonies never become large. Surface colonies are usually less than 2 mm. They are circular with uniform outline and a well-defined border, are grayish by reflected and bluish by transmitted light, are smooth and convex, are moist and glistening in ap- pearance, and are friable in consistency. "When magnified the surface colonies appear light brown in color, and granular in structure, the density decreasing from the center to the periphery. Deep colonies appear dense, dark brown, and coarsely granular. They are in general lenticular to oval but are sometimes almost spherical in form. EUROPEAN FOULBROOD. 13 Glucose gelatin plate. — At refrigerator temperature aud within 3 days, the surface colonies begin to liquefy the gelatin, each liquefied area appearing somewhat as a minute drop of water. Agar slant. — In one day numerous gray colonies cover the inoculated surface. Bouillon. — Within a day the medium is uniformly and moderately clouded. Fermentation. — In glucose, lactose, saccharose, levulose, maltose, and man- nite bouillons, a uniform clouding of the media occurs. The growth takes place in both arms of the tube, but is heavier in the open one. Considerable acidity, but no gas, is produced. Milk. — Milk is rapidly coagulated. Digestion of the coagulum follows. In from 3 to 5 days more than one-half has been changed. Within 24 hours the color is discharged in litmus milk, except at the top of the medium. In other respects it is like the plain milk. Potato. — No visible growth. That growth in the potato water takes place is confirmed by microscopic examination. Gelatin stab. — Liquefaction along the line of puncture is appreciable after one day. In four days a cylinder of liquefied gelatin 1 cm. in diameter surrounds the original line of puncture and soon extends to the walls of the tube. Pathogenesis. — No disease results when the brood of bees is fed cultures of Streptococcus apis either by the direct or indirect method. A rabbit and two guinea pigs inoculated with a pure cul- ture of Streptococcus apis were not sus- ceptible to infection with the species. BACTERIUM EURYDICE The presence of this species in European foulbrood was pointed out by the writer in an earlier pub- lication (15). Among the second- ary invaders in larvae infected with Bacillus plwton, Bacterium eury- ^'«- 4^--S'^<^^rium euryMce. dice is one of the earliest to be found. It is often present in consid- erable numbers. In plating for the species the stomach contents from larvse sick, but not dead, of the disease should be used. In studying this species cultures were isolated which in some respects differed from it. Whether these are different species or belong to a group of which Bacterium eurydlce is a representative has not been definitely determined. To isolate Bacterium eurydlce the plating has been done with glu- cose agar. Incubation must be carried out at room temperature. Growth of the species is always slow and never luxuriant. Under favorable conditions colonies are visible after one day. To preserve cultures they must be renewed frequently. Occurrence. — Bacterium eurydice is frequently present in larvse sick or recently dead of European foulbrood. Glucose agar plate. — To the naked eye the surface colonies are slightly convex, smooth, and glistening. They are from 1 to 2 mm. in diameter, cir- k 14 BULLETIN 810, U. S. DEPARTMENT OF AGRICULTURE. cular and uuiform in outline. The color is bluish by transmitted and grayish by reflected light. Under a two-thirds objective they are a light brown, and are finely granular near the periphery, but more coarsely granular near the center. Morphology.— The rod (fig. 4; PI. VII, C) is small and slender with slightly rounded ends, occurring usually in pairs or singly. It is nonmotile and no spores are produced. StaiHing properties. — It is stained easily and uniformly with the ordinary aniline sta1n.r in the seed left in the open field, evidently in the larval stage. ^ Figures in parentheses refer to " Literature cited," p. 20. CLOVER AND ALFALFA SEED CHALCIS-FLY. 6 DISTRIBUTION. In 1904 Dr. E. G. Titus (5) gave the following distribution of B. funehris by States: Hanford, Cal.; Fort Collins, Colo.; Marengo and Urbana, 111.; Winona Lake, Ind. ; Richmond, Kans. ; Agricul- tural College, Mich. ; Agricultural College, Miss. ; St. Anthony Park, Minn.; Quaker Street, N. Y. ; Corvallis, Oreg.; Providence, R. I.; Burlington, Vt. ; Pullman, Wash. ; Danville and Virginia Beach, Va., and the District of Columbia. Mr. George I. Eeeves reported this species from Lincoln, Nebr., the same year, and in 1915 Prof. F. M. Webster reported its presence at Tyngsboro, Mass., Kensington, Md., Lebanon, N. H., and Chambersburg, Pa. It was further reported from Wakeman, Ohio, in 1905, by W. B. Hall; at Albuquerque, N. Mex.^ 1908, Sacaton, Ariz., 1909, and central Utah, 1910, by C. N. Ainslie; at Piano, Tex., in 1909, by T. D. Urbahns ; at Twin Falls, Idaho, in 1911, by H. T. Osborn ; at Nashville, Tenn., in 1910, by George G. Ainslie; at Newell, S. Dak., in 1913, by C. N. Ainslie; at Alva, Okla., in 1913, by E. G. Kelly ; at Charleston, Mo., in 1914, by G. AV. Barber; at Minden, Iowa, in 1915, by T. D. Urbahns ; at Hagerstown, Md., in 191G, by H. L. Parker ; and at Fallon, Nev., in 1917, by T. D. Urbahns. LTnder later dates, B. funehris was reared from many additional localities by the writer and other persons connected with the Bureau of Entomology, so that at the present time it is known to be present in practically every locality in the United States where either red clover or alfalfa seed is grown to any extent. . In 1913 the writer found it in alfalfa seed reported by a seed dealer to have arrived from Germany. This species has also been reported from other foreign countries as follows: 1906, N. E. Hansen, Medicago falcata, Omsk, Siberia. 1907, C. V. Piper, alfalfa, Turkestan, Asia. 1908, Mr. Brand, alfalfa, Pisa and Matilla, Chile. 1908, N. E. Hansen, alfalfa, Tashkent, Turkestan, Asia. 1908, J. G. Sanders, alfalfa, Tashkent, Turkestan, Asia. 1910, C. P. Lounsbury, alfalfa, Cape Town, South Africa. CHARACTER OF ATTACK. The injury caused by the chalcis-fly consists entirely of the hollow- ing out of the developing seeds (PL II, fig. 2). By the time the alfalfa seed pods and the clover heads have matured the destructive work of this minute larva has been completed within the infested seed, and the result is a hollow seed containing the insect in one of its stages, or simply the seed shell from which the adult insect has emerged. (PI. II, figs. 3 and 4.) This pest in no way interferes with the growing of either alfalfa or red clover for forage purposes, except in so far as a very poor stand may be expected when infested and hollowed-out seeds are planted. The loss caused by this pest 4 BULLETIN 812, U. S. DEPAETMENT OF AGRICULTURE. is to the growing of alfalfa or clover for a seed crop, and the insect may therefore be considered strictly a seed pest. FOOD PLANTS. The following plants have been found to be food plants of the chalcis-fly, B. fwnehris, seeds of each of these having been found in- fested by this species: Alfalfa (Medicago satira). Bed clover (TrifoUum pratense). Bur clover {Medicago Mspida nigra). Bur clover (Medicago Mspida terebellmn). Bur clover (Medicago Mspida denticulata) . Medicago falcata. Medicago rut hernia. Medicago tunetanu. Medicago tvberculata. Medicago arabica. The seeds of the following list of closely related plants apparently are not attacked by the chalcis-fly: White clover (TrifoUum repens). Alsike clover (TrifoUum hybrid um). Yellow sweet clover (MeUlotns offlcinaJis). White sweet clover (Melilotus alba). Sour clover (MeUlotus indica). ECONOMIC IMPORTANCE. Table I shows the destruction of alfalfa and clover seed as found in samples of seed taken in the given localities. Table I. — Percentage of alfalfa and red cloi'er seed found infested by Brucho- phagus funebris. Localitv. Date of collection. Seeds examined. Total. Infested. 8S6 7 902 45 156 81 1,.374 723 1,000 361 1,100 260 1,053 222 1,236 400 • 1,091 291 1,005 268 1,115 315 1,228 311 1,255 672 1,364 489 1,125 245 1,028 533 1,157 330 872 747 691 212 263 217 1,307 50 866 237 3,007 1,761 263 83 574 195 Seed de- stroyed. Remarks. Aberdeen, Idaho. Do Albany, Oreg Bard,CaUf Do Do Do Do Do Do Do Do Do Do Do Do Do Do nishop, Calif. . .. Hiickeyc, Ariz... Hurley, Idalio. . Caldwell, Idaho. Chico, Calif Do Chino, Calif Sept. 19,1914 do Sept. 5,1914 May 7, 1915 May 12,1915 May 20,1915 May 28,1915 June 8, 1915 June 17,1915 June 23,1915 July 2,1915 do July 28,1915 Aug. 4,1915 do Aug. 8,1915 do Oct. 12,1915 Nov. 10,1914 Sept. 2,1912 Sept. 11,1914 Sept. 10,1914 Apr. 12,1912 Oct. 25,1914 Sept. 24,1912 Per cent. 0.7 5 52 52.6 36.1 23.6 21 32.3 26.5 26.6 28.2 25.3 54.4 36 21.7 48.1 28 85 30.7 82 3.8 27.3 5S.5 31 34 Alfalfa seed. Do. Red clover seed. Alfalfa seed. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. CLOVER AND ALFALFA SEED CHALCIS-FLY. 5 Table I. — Percentage of alfaJfn atnl red clover seed found infested hy Briiclio- phagus funebris — Continued. Locality. Date of collection. Seeds examined . Total. Infested. Seed de- stroyed. Columbus, Ohio Corcoran , Cali f Dos Palos, Calif Glendale, Calif Do Gunnison, Utah Do Do Do Do Logan, Utah Manti, Utah Mesilla Park, N. Mex. Do Murrietta, Cali f Myton, Utah Nephi, Utah Do Do Do Do Pendleton, Oreg Red Bluff, Calif Riverside, Calif St. Anthony, Idaho. . Salt Lake City, Utah. Do Do Do Do Do Do Somerton, Ariz. Tempe, Ariz Do Do Do Do Do Do Do Do Do Do Twin Falls, Idaho. Tulare, Calif Do Washington, D. C . Wellington, Kans.. Do Woodland, Calif... Yuma, Ariz Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Oct. 17,190<< Oct. 3,1912 Sept. 15,1913 Aug. 2fi,1912 Jan. 12,1914 Sept. 23, 1914 do do do do Dec. 28,191") Sept. 23, 1914 Aug. 30,1908 .....do Aug. 7,1915 Apr. 5,1912 Sept. 24, 1914 do do do do Sept. 9,1914 Sept. 3,iJ14 Sept. 17,1914 May —,1910 do do Oct. 17,1910 Nov. 22,1910 Sept. 20,1911 Jan. 16,1914 May 10,1915 Jmie 28,1912 July 1,1912 July 8,1912 July 15,1912 Aug. 10,1912 Oct. 29,1912 July 22,1913 do Aug. 20, 1^913 July 24,1914 Aug. 15,1914 Sept. 14, 1914 Oct. 1, 1912 Nov. 18, 1913 Oct. 20,1910 June 3, 1910 June 6, 1910 Sept. 2,1914 Aug. A up fi, 1913 7, 1913 do Oct. —1913 Dec. 19,1913 Dec. 22,1913 June 18,1914 do do do do Apr. 24,1915 June 11,1915 July 21,1915 .....do 559 317 993 311 603 816 876 835 297 591 1,527 1,097 561 531 1,140 1,000 994 845 700 1,046 935 764 966 492 629 100 400 500 500 107 450 1,387 1,235 167 348 670 684 301 321 899 2,270 939 1,445 1,185 1,117 600 655 361 400 600 558 1,143 833 2,174 915 484 1,377 335 1,076 391 585 554 706 1,047 1,295 1,106 50 118 264 32 145 34 105 59 11 35 680 92 130 170 177 395 35 119 131 19 23 229 147 109 23 37 149 186 286 77 5 52 664 Per cent. 8.9 37 26.5 12 24 4 12 7 3.7 5.9 44.5 8.4 30.5 47.1 15.6 39.5 3.5 13.5 18.5 1.8 2.4 30 1.5.2 22 3.6 37 62.8 37.2 57 72 1 3.7 53.8 237 403 235 134 412 778 439 408 212 322 154 243 96 333 480 98 109 277 .524 646 443 1,017 76 171 82 24 95 277 311 214 211 35.4 50 7:i 42 46 35 47 28 18 29 25.5 37.5 26 83 80 18. 9.5 24.4 24 72.5 91 74 22.7 16 21 4.2 17 39.2 29.5 16.5 19.1 Red clover. ' Alfalfa seed. Do. Do. Do. Do. Do. Do. Do. Do. Do.« Do. Do. 8 Do. Do. Do. 4 Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. a Do. 3 Do.s Do. 3 Do. 3 Clover. * Alfalfa seed. Do. Do. 5 Do. 6 Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Clover. 1 Alfelfa seed, « Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. 1 Note made by T. H. Parks. 2 Note made by E. L. Barrett. 3 Note made by C. N. Ainslie. * Note made by H. T. Osbom. 6 Note made by R. N. Wilson. « Note made by E. G. Kelly. The loss to the alfalfa-seed crop depends upon the total yield and upon the percentage of infestation. 6 BULLETIN 812, U. S. DEPAETMENT OF AGRICULTURE. This loss frequently varies from 50 to ttOO pounds per acre as a result of the destructive work of this pest. At an average price of 12 cents per pound the loss may then be estimated at from $6 to $48 per acre annually. Wliile this loss in the aggregate is enormous, there is frequently an additional loss which is caused by the planting of uncleaned seed, resulting in a poor stand and the loss of time and money in replanting. COMMON NAMES. Bruchophagus funehrls is commonly known throughout the East- ern States as the clover-seed chalcis-fly. Throughout the alfalfa seed-growing districts, which are located almost entirely in the Western States, this insect is better known as the alfalfa-seed chalcis fly. Many farmers, however, frequently refer to its destructive work as that of a seed weevil, while others simply refer to the injured seed as " blighted seed," and apparently are not familiar with the cause of this condition. MEANS OF DISPERSION. At the present time it is impossible to study in detail the means by which B. fwnehHs became distributed over the different parts of the United States, for the reason that it is now found in almost every locality where either alfalfa or red-clover seeds are grown to any extent. It is, in fact, very probable that this species is a native of this country. The shipment of infested seed from one locality to another has, without doubt, been a great factor in the dispersion of the alfalfa- seed chalcis-fly, since one can visit seed houses and frequently find hundreds of the adults present in the sacks of either clover or alfalfa seed. Their spread probably was greatly aided also by the early settlers, who traveled across the country with horses and w^agons, taking with them either clover or alfalfa hay bearing a few infested seed pods. Throughout the mountain States there frequently is found a chain of small alfalfa patches along the mountain streams connecting the larger irrigation valleys, where alfalfa plants, along the edges of these fields, develop seed pods, and render conditions favorable for the spread of this insect from one field to another. Infested seeds and pods are undoubtedly washed down the mountain streams, enter irrigation canals, and may become distributed with the water over new fields many miles away. The wind is an important factor in the dispersion of the adults of B. funehrls over adjoining fields. The writer has frequently ob- served lai-ge numbers of chalcis-flies being carried by the strong summer breeze from one field to another. On one occasion more than Bui. 812, U. S. Dept. of Agriculture. PLATE I. a The Clover and Alfalfa Seed Chalcis-Fly (Bruchophagus funebris). a. Adult; &, larva; c, pupa. Bui. 812, U. S. Dept. of Agriculture. Plate 1 1. The Clover and Alfalfa Seed Chalcis-Fly. A, Egg; B, larva in hollow seed; C, cluster of socd pods showing openings from which adults emerged" L), seeds hollowed by Bruchophagus funebrls. ' CLOVER AND ALFALFA SEED CHALCIS-FLY. 7 two hundred adults of B. funehris^ carried by the wind from a near-by field, alighted upon the writer's shirt in the course of a few minutes. Further observations revealed thousands of these chalcis- flies being carried by the winds. LIFE HISTORY. METHOD OF STUDY. The life history of this seed chalcis-fly was studied by selecting old alfalfa plants and first carefully removing all seed pods which had begun to develop. The fresh blossoms were then artificially pollenized and covered with thin cheesecloth bags. About one week later the bags were removed and a mica breeding cage was placed over the soft green seed pods. One or more adults of B. funehi^is were then placed in the cage for oviposition into the green seeds. The mica cage, with the adults of B. funehrls, was usually removed after one day, and the seed pods were again covered with cheesecloth bags. Here they were allowed to remain until desired for dissection and study of the various stages of the chalcis-fly, or until the first adults emerged. It was found that larvse of B. funebns^ which had been re- moved from their natural cavities in the infested seed, could be transferred to a cavity made in a small cork, and if properly cov- ered with a medical capsule, or a small glass vial, they could be reared to the adult stage without difficulty in the laboratory. The hibernating larvae and pupa? also were carried successfully through to the adult stage in cavities made between two layers of sheet cork. Examinations of matured seeds were made by soaking the seeds for a few hours and then dissecting them under a binocular micro- scope. The period of incubation was studied by allowing adults to ovi- posit into seed pods covered with mica cages and then dissecting some of the seed from day to day until the first eggs were found to have hatched. THE EGG. The egg (PI. II, A) of B. funehris is small, elongated, measuring about 0.2 mm. long and 0.08 mm. in thickness. It consists of a delicate membranous bag filled with a semiliquid substance. One end of the Qgg is terminated in a flexible point, while the opposite end is marked by a slender tube-like film measuring about 0.2 mm. in length. LARVA. The larva (PI. I, B) is grublike in shape. It averages 1.9 mm. long and 0.9 mm. wide when fully developed. The pointed brown mandibles are chitinous and visible to the unaided eye. With the ex- ception of the mandibles the larva is white. It has no feet or legs 8 BULLETIX 812, V. S. DEPARTMENT OF AGRICULTUEE. and is almost free from pubescence. The 13 body segments are at times quite prominent. PUPA. The pupa (Ph I, C) measures 1.8 mm. long. It is at first white, later the eyes and ocelli are brown, and before becoming adult the pupa turns black. All appendages are folded close to the body, and a thin pupal skin covers the insect in this stage of its development. ADULT. The adult (PL I, A) of 7>. fimehris is a very small black insect measuring about 2 mm. in length. Frequently thousands are seen flying over the horses and mower when the alfalfa seed crop is being harvested. They also collect in gi-eat swarms on the shady sides of alfalfa seed stacks, from which these insects emerge. Farmers fre- quently mistake the adults of this species for gnats, which are about the same size and are frequently very troublesome at the time when the chalcis-fly is most abundant. . SEASONAL HISTORY. EARLY DEVELOPMENT. The first development of the chalcis-fly may be found in February and March throughout the Southwest, when the hibernating larvse begin to transform to the pupal stage. Spring temperatures and moisture conditions determine the time of this transformation. In colder climates it is much later, and under desert conditions the larval stage may be prolonged indefinitely. FIRST APPEARANCE OF ADULTS. The first adults apjjear in the spring, about four or five weeks after warm weather has set in. It is probable that in southern California adults may be active in small numbers throughout the entire winter. Adults of B. fimehns were taken in the fields in the San Fernando Valley, Calif., as early as March 3, but they usually do not appear in large numbers until about April. At Yuma, Ariz., they make their appearance in March, and the first seed pods, which usually develop along the check ridges, become infested by the middle of April. On April 17 green pods were taken at the edge of a field at Yuma, and upon careful examination, 8 seeds showed very small larvse, 5 showed half-grown larvse, 15 showed full-grown larvsB, and 20 showed B. funehris pupse. E. G. Kelly observed adults of B. fune- hr'is in the field at Wellington, Kans., as early as April 21. H. T. Osborn reared them in out-of-door breeding cages in the same locality as early as April 17. In the valleys of central California the adults of B. funehris become active in April and oviposit into the seed pods of bur clover {Medicago hispida), which forms seed pods much earlier than alfalfa. CLOVER AND ALFALFA SEED CHALCIS-FLY. 9 Mr. Kelly reported adults of B. funehris active and ovipositing in alfalfa seed pods at Wellington, Kans., by June 10, 1911. OVIPOSITION. The adults of Bruchopliagus funehris appearing in early spring are not always able to locate developing seeds of their host plants at the time of their emergence, and it may be three or four weeks be- fore an opportunity for oviposition is found. In midsummer ovi- position takes place within a few days after the adult emerges. Green and half -grown seed pods are favored for oviposition. The female adult takes a position directly over the slight enlargement on the pod caused by the developing seed. The ovipositor is forced through the gi'een seed pod and into the soft watery seed, where the egg is placed. The time required for oviposition is about one minute. POSITION OF EGG. The egg is usually placed just beneath the inner integument of the seed, sometimes between the cotyledons, and frequently within the semiliquid contents of a cotyledon. It is placed singly, although oviposition into the same seed may be repeated. The mark left by oviposition may sometimes be seen in the form of a very minute spot on the surface of the seed, but this is usually not discernible even with the aid of a microscope. The relative position of the seeds in a pod in no way seems to influence the chances of infesta- tion. EGG STAGE. Observations made during the month of April at Pasadena, Calif., showed the period of incubation as requiring from 7 to 12 days. In the month of June the incubation period was found to be about five days. E. L. Barrett, assisting the writer, found eggs of B. funehris hatching in four days during very warm weather. LARVAL STAGE. The larva, upon hatching from the Qgg, finds itself surrounded with the soft contents of the growing seed. The first feeding takes place after the larva is one or two days old, and in midsummer rapid growth follows. The larva of B. funehris never leaves the seed in which it is developing to infest another seed in the same pod. Where two or more larvae chance to be in the same seed, one or both usually perish before development is completed. Molts, apparently, occur indefinitely in the larval stage, and appear to be the peeling off, in small fragments, of the old larval skin. The 136601°— 20— Bull. 812 2 10 BULLETIIs^ 812, U. S. DEPARTMENT OF AGRICULTURE. time required for the complete development of the larval stage va- ries under differing temperature conditions, and depends upon the condition of thci seed in which it develops. Under the most favor- able conditions, as observed by the writer, the larvae completed de- velopment in 12 days. In early spring this period was prolonged to 30 days. PERIOD OF AESTIVATION. The factors of greatest influence in the life cycle of B. funehris are those which determine tlie period of aestivation. Larva? of this species, reaching their full larval development in soft green seeds, at once transform to the pupal stage and soon emerge as adults. On the other hand, if the infested seeds mature and harden before the larval development is fully completed aestivation follows. This resting period frequently continues throughout the remainder of the season, or until conditions become favorable for pupation and trans- formation to the adult stage. Atmospheric humidity, temperature, and irrigation are important factors determining the duration of aestiA'ation. Aestivation occurs frequently in the larvaj of />. funebris during the months of July, August, and September. Favorable conditions may at any time cause transformation to the pupal stage oi" aestivation may continue into hibernation. The transformation usually occurs at any time between March and July of the following season, though in ex- ceptional cases the larval stage may be continued into the second year. Alfalfa seeds collected in the fields on September 24, 1912, and kept dry in the laboratory showed adults of B. funehris emerging thi'oughout the season of 1913 and larvae living in the seeds until the spring of 1914. The last adults of B. funehris emerged as late as September, 1914. This habit causes great diffusion in the adults of the species. It was found that some of the progeny of a single adult developed and reached maturity in about 30 days, while other progeny from the same adult and from eggs deposited on the sam-^ day did not mature until the next year. PUPATION. The period of pupation in early spring requires an average of about 30 days, as shown by Table II. As the season advances and the days become warmer the duration of the pupal period is nmch shortened. Pupa' developing in midsummer have a pupal period of from fi to 10 days, and under the most favorable conditions this period is without doubt shorter than that observed in the breeding cages. , CLOVER ATTD ALFALFA SEED CHALCIS-FLY. 11 LATE APPEARANCE OF ADULTS. While some of the larvae apparently go into hibernation as early as August, adults may be seen emerging from the infested seeds and seed pods in the months of October and November. At Glendale, Calif., adults were still active in the fields as late as November 18, in 1912. C. N. Ainslie reported B. funehris as being numerous among the florets of red clover on September 17, in 1913, at Fairmont, Minn. At Albany, Oreg., the writer found them active over red clover on September 15, in 1914. At Nephi, Utah, they were active on alfalfa florets on September 24 of the same year. Activity of the adults, oviposition, and development of larvae usually continue as long as green seed pods are present in the fields. The first heavy frost kills the adults and pupae and most of the undeveloped larvae. The writer's observations show that in southwestern Arizona this species may have as many as four generations in a single season. This is, of course, under favorable conditions — first on volunteer alfalfa and then in the irrigated seed fields. Without doubt it com- pletes at least two generations in most of the intermountain seed- growing districts, where tlie growing seasons are frequently very short. The minimum number of generations in a single season may be less than one under very unfavorable conditions, as it required more than one season for some of the individuals under observation to reach the adult stage. Table II. — Lengtli of the pupal period for 50 indiriduals of Bruehophagua fune- hris which hibernated in the larval stage. Pupal period of male. Days. Mar. 10 to Apr. 14 35 Mar. 11 to Apr. 22 42 Mar. 10 to A pr. 21 42 Feb. 19 to Apr. 12 52 Mar. 10 to Apr. 22 33 Mar. 14to Apr. 6 23 Mar. 17 to Apr. 16 30 Mar. 1 1 to Apr. 2 22 Mar. II to Apr. 4 24 Mar. 14 to Apr. 6 23 Mar. 14 to Apr. 10 27 Mar. 14 to Apr. 11 28 Mar. 23 to Apr. 2.5 23 Mar. 18 to Apr. 19 32 Mar. 11 to Apr. 1 21 Mar. 11 to Apr. 4 24 Mar. 11 to Apr. 4 24 Mar. 16 to Apr. 13 28 Mar. 14 to Apr. 4 21 Mar. 17 to Apr. 15 29 Mar. 13to Apr. 10 28 Mar. 18 to Apr. 14 27 Mar. 11 to Apr. 2 22 Mar. 11 to Apr. 6 26 Mar. 11 to Apr. 2 22 Average 28. 3 Pupal i)eriod of female. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. Mar. 10 to Apr. 22 43 10 to Apr. 14 35 11 to Apr. 14 34 10 to Apr. 22 43 11 to Apr. 22 42 11 to Apr. 21 41 20 to Apr. 28 39 11 to Apr. 14 34 21 to Apr. 24 34 10 to Apr. 14 35 17 to Apr. 22 36 10 to Apr. 14 35 10 to Apr. 14 35 10 to Apr. 22 43 14 to Apr. 13 30 16 to Apr. 13 28 18 to Apr. 18 31 18 to Apr. 19 32 12 to Apr. 21 40 11 to Apr. 6 26 14 to Apr. 11 28 14 to Apr. 15 32 14 to Apr. 11 28 13 to Apr. 15 33 11 to Apr. 6 26 Average 34. 5 12 BinXETIN 812, XJ. S. DEPAKTMENT OF AGRICULTUEE. HIBERNATION. B. funehris hibernates in the larval stage within the infested seed, and the first larvae enter hibernation as early as August. In fact, many larvae, after spending a part of the summer in aestivating, continue into hibernation without resuming apparent activity. On the other hand, large numbers of larvae of this species can be found active in the seeds throughout September and October, and a few of them have been found feeding in the green alfalfa seeds throughout the month of December and even until they were killed by heavy frosts in early January. Early spring rains, together with warm days, mark the end of hibernation for a large percentage of the larvae. ! 1 \ \ \ 1 1 1 f X 1 X 1 \ \ \ \ 1 ^ 1 ( V 1 ^ X ^ "< —z =- =- """ = — '=z - - : p- ■^ "1 = : z. ■"" - _ = C=3 a =. Fig. 1. — Diagi-am showing development of Bruchophagus funebris. Each heavy line shows the time required for one individual from the egg to the adult stage. In southern California the first pupae may be found in the latter part of February, and these increase in great numbers throughout the month of March. H. T. Osborn, formerly of the Bureau of Entomology, reported B. funehris entering the pupal stage at Wel- lington, Kans., during the month of March under conditions slightly unfavorable to pupation. Some of the individuals continue to rest in the larval stage until late in the spring, and frequently until mid- summer. The greatest number of hibernating larvae are found along the edges of the field, along fence lines, and in neglected seed fields. In fact, they may be found wherever large numbers of in- fested seeds have been allowed to remain on the fields unharvested. In figure 1 each heavy line shows the time required for the de- velopment of a single individual of B. fimehns. The different lengths of these lines show the period of development from the egg CLOVER AND ALFALFA SEED CHALCIS-FLY. 13 to the adult stage to be from 26 days to 24 months, and, in a few cases, to carry the individuals through two winters before the adult stage is reached. RELATIVE INFESTATION THROUGHOUT THE SEASON. During the season of 1915 newly formed alfalfa seed pods were taken at Bard, Calif., at intervals of about one week, and examined for infestation by the seed chalcis-fly. It was found that the seed pods taken as early as April 24 showed 39 per cent of the seeds to be infested. On May 8, as many as 52 per cent showed infestation by the chalcis-fly. This was undoubtedly due to the increased emer- gence of the chalcis-fl3\ The period from the middle of May to the niiddle of July, at which time the pods were forming in gi'eat Alfalfa along fence lines, ditch banks and Wa^-fe areas Alfalfa seed fields Mo. ^93^ Lor-vae Pupae Adults Effgs Lar vae Pupae Adults Feeding Resting reeding f?es ting /irh. ■^■m' ^ m Mar. .^^ ^^ B 1 ^ Apr- » \ ^^ "^^ ^ " Ma^. ^ ^" "^a^ H^^ k B Jun ^^ ^ ^^ ^^^ ^A^ ^ ^h. Aug. Sep. Oct Un. — , ^^^ \ ^_ ^^^ Mi^M ^^^ k a^^ ^= =^ — ^ ^I ^^ — sr — „^ — " ■ ^ ^r~ .^ ^ ^^ ^^ u j^"" Dec ^ "" ^^_ r 1 ~ Jan. ' ~ Feh. M Fig. 2. — Diagram showing comparative abundance of hris during the different months the stages of Bruchophagus fune- of the year. abundance, showed the infestation dropping to an average of 28 per cent. The average infestation for the period from the middle of July to the end of September increased to 49.5 per cent. This was un- doubtedly due to the increased number of the chalcis-flies and to the rapid decrease in the seed pods caused by the cutting of alfalfa on many of the seed fields. By October all the seed fields had been harvested, and the chalcis- flies were forced to breed in the green pods still forming along the edges of the fields. Seed pods collected on October 12 showed 85 per cent of the seeds to be infested. Figure 2 shows the comparative abundance of the stages of B. funehris during the different months of the year. It will be noted by the chart that eggs of B. funehris are found earlier along fence lines than in the seed field. This is due to the 14 bulleti:n^ 812, u. s. depaetment of agriculture. fact that isolated plants on wast« areas form the earliest pods, and the seed fields are cut for hay before the seed crop is grown. A sudden scarcity of eggs, feeding larvae, and pupae is seen in the field in October, wliich is due to the ham^esting of the latest seed crops. Eesting larvse continue to remain on the field throughout the winter in remaining seed pods, but are much more abundant along fence lines. Spring irrigation of seed fields hastens emergence of adults, and this results in the absence of resting larvae and pupse on most fields during the month of May, PARTHENOGENESIS. The parthenogenetic habit is well established in the females of this species. Female adults, reared from pupae which were com- pletely isolated, were placed in cages for oviposition into green alfalfa seed pods. All precautions were taken against earlier in- festation of the green seeds. These virgin females oviposited freely, and their progeny was reared in breeding cages. The largest num- ber of offspring from a single female observed by the writer was 20, and all of the progeny of these virgin females proved to be males. CONTROL METHODS. The practical methods of controlling the chalcis-fly in the alfalfa and clovefr seed fields are mostly cultural methods. It can not be hoped to eradicate this pest from any seed-producing locality; but with the proper methods effectively applied thousands of dollars can be saved annually to the different seed-growing districts. Be- cause of the rapid dispersion of the adults of this species it is of vital importance that control methods should be taken up by com- munity action rather than by the individual. BURNING OVER OF FENCE LINES. Everywhere in the alfalfa and red clover seed-growing districts one! will find large quantities of standing plants along the fence lines and waste areas bearing the seeds infested by hibernating larvae. Many hibernating larvae of the chalcis-fly can be destroyed in the fall of the year by burning over such areas, thereby greatly lessen- ing the abundant emergence of adults in early spring. It has been found that live stock can not be depended upon to rid the fence lines of standing alfalfa in the winter unless there is a great short- age of forage. WINTER CULTIVATION. The cultivation of alfalfa seed fields during the winter is of much value in covering the infested alfalfa seeds containing hibernating larvae. It is not necessary that infested seeds should be covered to Bui. 812, U. S. Dept. of Agriculture, Plate III. Tetrastichus bruchophagi. A, Larva; B, pupa; C, adult. Bui. 812, U. S. Dept. of Agriculture. PLATE IV. LlODONTOMERUS SECUNDUS. A, Adult female; B, larva; C, pupa. CLOVER AND ALFALFA SEED CHALCIS-FLY. 15 sucli a depth that the emerging adults are unable to reach the sur- face, because the slightest covering of moist soil will cause many of the infested seeds to mold, and prevent development of the pupal or adult stages. The necessary cultivation can be accomplished by the use of an alfalfa cultivator, spring-tooth harrow, or a disk. Care should be taken that the surface of the field is left smooth, to facilitate close cutting, as well as to prevent clods and gravel from being taken up with the seed when the seed crop is har^^estcd. IRRIGATION OF ALFALFA SEED FIELDS. Whenever possible, the alfalfa seed grower should do most of his irrigating in early spring, or before the seed crop is grown. Where this can be practiced, two important points may be gained. First, the tender, rank growth of alfalfa usually following irrigation will be removed as an early fodder crop, and a steady gi'owth of more substantial stems will take its place to produce the seed crop. Second, irrigation of a seed field stimulates the emergence of the chalcis- fiies which remain within the infested seeds upon the ground. The humidity of the atmosphere over an irrigated field accelerates the emergence of this pest from seeds of the newly forming crop, which would otherwise become sufficiently dry to force many of the larvse into a resting period. It has been supposed by farmers that flooding the seed fields after the first crop of alfalfa has been removed would destroy the larvae of the chalcis-fly present in alfalfa seeds upon the ground. The writer has found that soaking infested seed in water for four or five days is not sufficient to kill the larvae resting within the seeds. This would render any attempt at their destruction by irrigation im- practical. CUTTING EARLY PLANTS ON WASTE AREAS. The fence lines, ditch banks, and other waste areas where volun- teer alfalfa can be found, are among the great sources of infestation. It has been found that the chalcis-fly frequently passes through one or two generations on this volunteer alfalfa before the seed fields are advanced to the point where infestation takes place. Every farmer, therefore, in an alfalfa seed-growing district should cut all of the standing alfalfa along the fences and waste areas at the time a hay crop is cut before he grows a seed crop. Alfalfa seed pods should not be allowed to develop anywhere around the field until the plants of the regular seed crop have begun to form pods. CAREFUL CUTTING OF HAY CROPS. It is of the greatest importance that special care should be exer- cised in cutting alfalfa or red clover before the seed crop is grown. Ik- 16 BULLETIX 812, U. S. DEPAETMENT OF AGRICULTURE. Any plants carelessl}' cut, so that some stems remain standing, and allowed to develop seed in advance of the regular seed crop, offer favorable conditions for the multiplication of the chalcis-flies. All pocket gophers should be poisoned or trapped and their mounds, as well as other irregularities in the surface of the field, should be leveled so as to make clean cutting possible. PASTURING BEFORE GROWING SEED. After removing the first crop of alfalfa for hay, and preparing to grow a seed crop, it is frequently possible to secure a herd of live stock, which may be turned into the field for two or three days. If a sufficient number of horses, cattle, or sheep can be secured they will render a valuable service by destroying the clusters of plants which frequently escape the mower, develop early seed pods, and form favorable breeding places for the development of the chalcis- ily before the regular seed crop has become sufficiently advanced. ALLOWING THE SEED CROP TO STAND TOO LONG. The development of seed pods in the alfalfa seed fields is to a great extent periodical. That is to say, the first set of pods will have been formed when a new set of blossoms appears, or the first set of pods will be reaching maturity when the second set of pods is soft and green. This condition is most noticeable in fields receiving frequent irrigation. The tendency on the part of the farmer is to allow the seed crop to remain on the field too long, while waiting for a third or even a fourth set of pods to develop. It has frequently been ob- served that the chalcis-fly passes a complete generation in the earlier pods, and, in much greater numbers, infests the later pods forming on the same plants. Consequently, all seed fields should be handled, as much as possible, so that a heavy setting of pods will be secured with the first bloom and the crop removed from the field as early as possible. This will avoid a greater percentage of infested seeds as well as loss of time and will make room for the growing of subse- quent crops for hay. SECOND CROP OF SEED. In the Southwest, where there is little danger of early frost, seed growers occasionally attempt to grow two crops of alfalfa seed in a single season. The results of such instances are usually very disap- pointing, on account of the usual severe infestation common to the second crop. This is frequently so great that the crop does not pay the harvesting expenses. An attempt to grow a second crop of alfalfa seed in a single sea- son will meet with almost certain failure as a result of the increased abundance of the chalcis-flies late in the summer. Bui. 812, U. S. Dept. of Agriculture. Plate V. EUTELUS BRUCHOPHAGI. A ^dult female; B , larva; C, pupa. Bui. 812, U. S. Dept. of Agriculture. PLATE VI. LlODONTOMERUS PERPLEXUS. A, Adult female; B, larva; C, pupa. CLOVER AND ALFALFA SEED CHALCIS-FLY. 17 PASTURING INFESTED FIELDS. The great mistake of pasturing alfalfa seed jEields which have be- come too severely infested to yield a crop worth harvesting is too fre- fjuently made. Live stock walking through the ripe alfalfa cause the seeds to be shelled out and those infested with hibernating larvae of the chalcis-fly fall on the ground, where they become the source of infestation for the following year. HYMENOPTEROUS PARASITES. TETRASTICHUS BRUCHOPHAGI Gahan (6). Tetrastlchus hruchophagi^ (PL III) is generally distributed over most of the northern half of the United States. In the central Cali- fornia alfalfa seed-growing sections it is the most active of the pres- ent known parasites of BrucK.ofhagus fwnehris. While it can not be depended upon to control the alfalfa-seed chalcis-fly, it has been ob- served to destroy about 50 per cent of the larvae of the chalcis-fly normally in the fields. This species develops within the seed and feeds upon the larval stage of its host. LIODONTOMERUS SECUNDUS Gahan (8). LiodontOTnerus secvrndus ^ (PL lY) has been found by the writer to be an active parasite upon the chalcis-fly when the latter infests the seeds of red clover. It does not seem to attack the chalcis-fly larvse infesting alfalfa seeds. This species is apparently most active in the red-clover seed-growing sections of Oregon and Idaho. It is active in the fields throughout the summer and hibernates in the lar- val stage within the infested seeds left on the field. EUTELUS BRUCHOPHAGI Gahan (8). Eutelus hi^chophagl- (PL V) is of economic importance as a parasite of the chalcis-fly in the mountain valleys of southern Idaho and central Utah. It has also been found by the writer east of the Sierra Nevada Mountains in northern California. It does not appear to be present in the southwestern alfalfa seed-growing districts. LIODONTOMERUS PERPLEXUS Gahan (8). Liodontomerus perplexus- (PL VI) is a parasite of considerable importance in checking the abundance of the chalcis-fly breeding in alfalfa seeds throughout western Arizona. It was reared as far north * For a more complete account of the species, see Urbalins (10). 2 For a more complete account of this species, see Urbahns (11). 18 BULLETIN 812. U. S. DEPAETMENT OF AGRICULTURE. as South Dakota by C. N. Ainslie. This parasite attacks the larvae of its host, and develops within the infested seeds. It has not been observed attacking the chalcis-fly when the latter infested seeds of red clover. This species is most active during midsummer. It ap- pears slower in resuming activity in sj^ring than the other species studied. HABROCYTUS MEDIC AGINIS Gahan (7). Habrocytus niedlcagbus ^ (PL VII) is present throughout Cali- fornia, Idaho, Utah, and parts of Arizona. It has also been reported from South Dakota, Kansas, New Mexico, Nevada, and Oregon. "While this parasite can not be depended upon at the present time to control the chalcis-fly in any locality where it has been observed, it is of value in checking the abundant development of its host. This species is parasitic externally ujDon the larvse of its host, and com- l^letes its development within the infested seed. From all observations, Habrocytus medicaginis attacks the chalcis- fly only when the latter infests the seeds of alfalfa. This parasite has never been found ui:>on the same host within red clover seeds. TETRASTICHUS VENUSTUS Gahan (7). T etrastichiis venustus does not occur in sufficient numbers to be considered of economic importance as a parasite of the chalcis-fly in alfalfa seeds. While this species has been reared in small numbers from alfalfa seeds infested with the seed chalcis-fly, and is without doubt a parasite of the latter, the larval stage has not been recog- nized. Specimens were reared by the writer from Corcoran, Red Bluff, and Tulare, Calif., and from Yuma, Ariz. TRIMEROMICRUS MACULATUS Gahan (6) .2 Trirrwromicrus "inaculatus (PL VIII) is of economic value as a para- site of the alfalfa-seed chalcis-fly throughout Arizona and Cali- fornia. It has been found especially active in the Buckeye and Yuma Valley seed districts of Arizona, and in the Honey Lake and Im- perial Valleys of California. LIODONTOMERUS INSUETUS Gahan (8). Notes on file in the Bureau of Entomology, made by H. T. Osborn, under date of September 19, 1910, show the larvse of this species to be parasitic upon larvae of Bmchophagm funebris. Speci- mens m the National Museum were reared by C. N. Ainslie, Mesilla ^ For a more complete account of this si>ecies, see Urbahns (9). = For a fuller description, see Urbahns (11). Bui. 812, U. S. Dept. of Agriculture. Plate VII. />-^ ,>-^ B L »i ^^ ^^'^ Habrocytus medicaginis. A, Adult; B, cages for rearing parasite larvae; C, larva; D, larva destroying its host larva; E, pupa. Bui. 812, U. S. Dept. of Agriculture. Plate VIII. Trimeromicrus maculatus. A, Adult female; B, lar\'a; C, pupa. CLOVER AND ALFALFA SEED CHALCIS-FLY. 19 Park, N". Mex.; H. T. Osborn and E. G. Kelly (1910), Wellington, Kans. ; E. G. Smyth and R. N. Wilson, Tempe, Ariz. ; and by E. G. Kelly, Sedgwick,'Kans. (1914). The economic importance of this species as a parasite of the chalcis-fly is not known at the present time. EUPELMUS sp. A single larva of this species was dissected from an alfalfa seed by M. Marshall, and was found to be parasitic upon the larva of B. funebiis infesting the seed. The specimen was taken at Pasadena, Calif. The larva of this species measured about l.G mm. in length, and was almost white in color. The head was small and the chitinous mandibles distinctly visible. The body segments w^ere conspicuous and covered with fine pubescence. The larva was rather active. The pupa, at first white, turned to a smoky-white color, and almost black before the emergence of the adult. PREDACIOUS MIDGE. LESTODIPLOSIS sp. Felt. Larvae of a midge, described as Lestodiplosis sp. by Dr. E. P. Felt, were found in infested alfalfa seeds, where they had apparently de- stroyed the larva of B. funehns. It is quite probable that larvae of this midge confine themselves to cracked or broken seeds, and where the larva of the chalcis-fly is exposed. The specimens under observa- tion were taken in San Diego County, Calif. LITERATURE CITED. n (1) COMSTOCK, J. H. 1880. Report of the Entomologist for 1879. In Kept. U. S. Commis- sioner of Agric, p. 185-373. (2) ASHMEAD, W. H. 1894. Descriptions of new parasitic Hymenoptera. In Trans. Amer. Ent. Soc, V. 21, p. 318-344. (3) HopKixs, A. D. 1896. Some notes on observations in West Virginia on farm, garden, and fruit insects. In Proc. 8tli Ann. Meeting of Assn. Econ. Ent. U. S. Dept. Agr., Div. Ent., Bui. 6, n. s., p. 71-74. (4) 1898. Some notes on observations in West Virginia. In Proc. 10th Ann. Meeting Assn. Econ. Ent. U. S. Dept. Agr., Div. Ent., Bui. 17, n. s., p. 44-49. (5) TiTXTs, E. S. G. 1904. Some preliminary notes on the clover-seed chalcis fly. In Some Miscell. Results of the Work of the Div. Ent. U. S. Dept. Agr., Div. Ent., Bui. 44, p. 77-80. (6) Gahan, a. B. 1914. New Hymenoptera from North American. In Proc. U. S. Nat. Mus., V. 46, p. 431-443, pi. 39. (7) 191.5. Descriptions of new genera and species, with notes on parasitic Hymenoptera. In Proc. U. S. Nat. Mus., v. 48, p. 163. (8) 1917. Descriptions of some new parasitic Hymenoptera. In Proc. U. S. Nat. Mus., V. 53, p. 195-217. May 26. (9) Urbahns, T. D. 1916. Life history of Habrocytus medicaginis, a recently described parasite of the chalcis fly in alfalfa seed. In V. S. Dept. Agr., Jour. Agr. Research, v. 7, no. 4, p. 147-154, pi. 4, 1 fig. (10) 1917. Tetrastichus bruchophagi. a recently described parasite of Bruchophagus funehris. In V. S. Dept. Agr., Jour. Agr. Re- search, V. 8, no. 7, p. 277-282, pi. 78. (11) 1919. Life-history observations on four recently described parasites of Bruchophagus funebris. In U. S. Dept. Agr., Jour. Agr. Re- search, V. 16. no. 6. pi. 165-174, pi. 22--23, 8 fig. 20 ADDITIONAL COPIES OF THIS PURUCATION MAY BE PROCURED FROM THE SUPERINTENDENT O? DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY V UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 826 Contribotion from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER August 10, 1920 GENERIC CLASSIFICATION OF THE HEMIPTEROUS FAMILY APHIDIDAE By A. C. BAKER, Entomologist Deciduous Fruit Insect Investigations CONTENTS Saperfamily Aphidoidea 2 Phylogenyof the Aphididae 3 Key to the Subfamilies of the Aphididae 10 Subfamily I, Aphidinae 10 Tribe Lachnini 12 Tribe Thelaxini 20 Tribe Calliptorini 21 Tribe Greenideini 37 Tribe Setaphidini 38 Tribe Aphldini 39 Subfamily n, Mindarinae 61 Page Subf&mily IH, Eriosomatinae 62 Tribe Eriosomatinl 65 Tribe PcinphiginI 68 Tribe Melpkini 73 Tribe ProciphilinI 75 Tribe Fordini 77 Subfamily IV, Hormaphldinae .... 81 Tribe Hormaphidinl 83 Tribe Oregmini 84 Tribe Cerataphidini 86 Genera Not Placed 88 WASHINGTON GOVERNMENT PRINTING OFnCB 1920 UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 826 Contribution from the Bareau of Entomoloey L. O. HOWARD, Chief O^'^i. ,y^*^Wd. Washington, D. C. PROFESSIONAL PAPER. August 10, 1920 GENERIC CLASSIFICATION OF THE HEMIPTEROUS FAMILY APHIDIDAE.' By A. 0. Baker, Entomologist, Deciduous Fruit Insect Investigations. CONTENTS. Page. Superfamily Aphidoidea 2 Phylogeny of the Aphididae 3 Key to the subfamilies of the Apiiididae 10 Subfamily I, Aphidinae 10 Tribe Lachnini 12 Tribe Thelaxini • 20 Tribe Callipterini 21 Tribe Greenideiui 37 Tribe Setaphidini 38 Tribe Aphidini 39 Subfamily II, Mindarinae , 61 Subfamily III, Eriosomatinae 62 Tribe Eriosomatini 65 Tribe Pemphigini 68 Tribe Melphiui 73 Tribe Prociphilini 75 Tribe Fordini 77 Subfamily IV , Hormaphidinae 81 Tribe Hormaphidinl 83 Tribe Oregmiui 84 Tribe Cerataphidini 86 Genera not placed 88 Probably no group of insects has received more attention at the hands of economic entomologists than aphids, or plant-lice. Their interesting and often complicated biologies have attracted the atten- tion of investigators, not only among entomologists, but among workers in the larger fields of zoology and general biology. While a large amount of work on the life histories and biologies of aphids has been done, corresponding progress in their classification has not been made. This is probably due to several causes, such as the lack of correlation of biologic and taxonomic facts, and the failure of aphidologists to consider sufficiently the results of the work of others. On account of the great economic importance of aphids, and the necessity of their study in the development of control measures, the lack of knowledge concerning their systematic relationships results .in much confusion. Some biologic workers, in fact; do not now at- tempt to give the name of the species being studied on account of the difficulty experienced in securing correct determinations. 1 This paper is the first of a series treating the Aphididae. It will be followed by others deaUng with the economic importance, biologies, and relationships of species in the different genera. 141613°— 20— Bull. 826 1 2 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. The present work was undertaken in the hope of remedying such a condition to some extent at least. The genera of the world have been studied. Many workers have lent material and the large col- lections of the National Museum and Bureau of Entomology have been drawn upon. In the National Museum collection a large per- centage of types has been available. To the study of preserved material have been added embryological, anatomical, and biological investigations that a better understanding of the natural relationships might be gained. Besides many aphidologists in this country and abroad, who have given helpful suggestions and many of whom have read and criticized the manuscript, the writer is indebted to Dr. A. L. Quaintance, of the Bureau of Entomology, for the facilities for conducting many of the biological investigations which to a large extent have laid the foundation for the systematic treatment here given. Superfamily APHIDOIDEA. There appear to be two distinct families in the superfamily Aphi- doidea. These are the Aphididae and the Phylloxeridae. The present paper deals only with the Aphididae. Members of the Phylloxeridae differ markedly from forms belong- ing to the Aphididae. In the first place their biologies are quite different in that parthenogenetic oviparous forms occur during the summer. In the Aphididae only the sexed females which are pro- duced in the fall are normally oviparous. In structure the two families are separated at once by the formation of the stigma of the forewing. The wing itself seems very little different in an Adelges or Phylloxera from that in some of the special- ized genera of the Aphididae. An examination of the freshly emerged wing, however, as has been pointed out by Dr. Patch, shows that the stigma in the ^Phylloxeridae is formed by the radial sector and the stigmal vein is the media. In the Aphididae, on the other hand, the stigma is formed by radius ^ and the stigmal vein is the radial sector. The two families may thus be separated as follows: Key to the Families of the Aphidoidea' Summer parthenogenetic oviparous forms produced: Stigma formed by the I'adial sector Phylloxeridae. Only sexual oviparous forms produced: Stigma formed by radius j Aphididae. A word of explanation in regard to the name Phylloxeridae may be necessary. The genus Chermes was erected by Linnaeus in 1758 and in 1862 was replaced by Psylla Geoff roy. For this genus Chermes Jicus L. was set as type by Lamarck in 1801. Ficus, therefore, be- comes i'pso facto the type of Chermes, and Chermidae the family name of the "jumping plant-Uce." The family name for the aphi- doidcan group, therefore, is derived from the genus Phylloxera Boyer (1834). GENERIC CLASSIFICATION- OF APHIDIDAE. 3 PHYLOGENY OF THE APHIDIDAE. In many published classifications of the Apliididae those groups which according to the writer's conception are tlie most specialized have been placed as the most primitive. This is the case with those TsmAA'^/^/' £/-/o sof7Ta,^/i ' Ce^a''u6A/d/if /¥or/n a^/i /(//n i ' Tlx //a. u/AV/zz a. 7^/n(fah//iae 'Penfa/on/na. T'^a./Tt/z^a^ Fig. 1.— Phylogeny of the hemipterous family Aphididae. insects forming in the present classification the Eriosomatinae and the Hormaphidinae. A study of the anatomy and the biology of aphids makes it evident that there are three main groups of living forms for which subfamily 4 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. names are here used: Aphidinae, Eriosomatinae, and Hormaphidinae. Besides these there is the one species, in some ways a relic of the past, forming the Mindarinae. As will be seen by the diagram (fig. 1), the Aphidinae is considered the most primitive subfamily of the three main groups. This is substantiated by the habits of the insects, by the structure of the wings, beak, etc., and by the sexual forms. Practically all of these msects are foliage or twig feeders. They live, as a rule, in colonies and have not developed any very special methods of life, such as highly complex gall formation. With the exception of the Mindarinae the most primitive wing structure oc- curring in the family is met with here. " The media is most commonly twice branched throughout the subfamily and even in the Mindarinae there is more of a reduction than this. The antennae are of six segments, the largest number found in the family, and the sen- soria are simpler in nature than those met with in the other two large subfamilies. The beak in the Lachnini shows also a primitive condition in its segmentation. The sexual forms are most primitive in the Aphidinae. Winged sexes often occur, at least the males are very commonly winged. Both sexes still retain their beaks and feed on their hosts in the same way as do the other forms, and the ovaries of the female develop normally, and she produces several eggs. When the phylogeny of this subfamily is studied, there becomes evident the primitive character retained by the Lachnini. In these forms the beak structure and the nature of the antennae and cornicles point to a primitive condition. The sexes, too, indicate this, though not markedly more than in other tribes. But the fact that these forms are mostly conifer feeders should not be overlooked. It is the opinion of the writer that this is a primitive habit. The Lachnus branch, therefore, may be considered the lowest branch of the Aphidinae. If the wings of fossil aphids be examined it will be seen that by far the greater number of them possess a wing structure quite different from that of our living forms. The radial sector arises back of the stigma, which is usually very long and narrow. This character is retained probably only in the Mindarinae. It is evident, then, that during the development of the present Aphidinae this vein migrated toward the tip of the wing until it came to stand either in the middle of the stigma or near its tip. On one line of this migration is the Lachnina wherein the vein has reached nearly to the tip of the wing and become short and straight. The remain- ing characters apart from the wings have in these forms remained quite primitive. The subtribe Eulachnina is evidently a more specialized group on this same line of development, for it possesses GENERIC CLASSIFICATIOlSr OF APHIDIDAE. 5 the same type of wing. Therefore this subtribe is considered quite closely related to the subtribe Laclinina but differing from it in specialization of body form, cornicles, and eyes. The other sub- tribes of the Lachnini have quite a different wing structure. While the radial sector has changed its position considerably from that found in the fossils it has not reached the tip of the stigma and is not straight, but much curved. In this regard, therefore, the Pterochlor- ina is perhaps the most primitive subtribe in the Lachnini, although in many respects it is specialized. On the other hand there are two highly specialized subtribes, the Anoecina and the Tramina. It is usually the custom to place the Anoecina with the Pemphigina. Its relations, however, are here. The adult forms are very similar in- deed to the lachnids. The sexual forms,, on the other hand, are small and apterous and suggestive of the sexes of the Eriosomatinae, and there is considerable ground for placing the Anoecina there. These sexual forms, however, seem to differ quite distinctly from those of the Eriosomatinae, which are beakless and the oviparous female of which never develops more than one egg. The develop- ment of the stigma shows quite an extreme modification from the long, narrow, primitive stigma. Near this same line of development is the rather highly specialized subtribe Tramina. The most marked character of this subtribe is the extreme modification of the hind tarsi. In considering only the genus Trama it might be thought that the tribe should belong with the Eriosomatinae. The species troglodytes has often been figured with cornicles in the apterous form. Specimens from Mordwilko and Schouteden determined as this species lack them and the writer therefore considers Trama as the most specialized genus in the sub- tribe. Another genus represented by radicis Kalt. shows cornicles very large and of a typical Lachnus character. Through this genus, therefore, the subtribe can be placed at once with its relatives in the Lachnini. Apart from the peculiar tibial character this genus is very lachnid-like. The next branch from the- Aphidinae is the Callipterus branch, which may be considered as arising somewhat later than the Lachnus branch. From this offshoot soon after it arose and before the present genera of the Callipterini appeared the Thelaxini separated. This tribe, the old Vacuini, also has usually been placed with the Eriosomatinae. There are some resemblances, it is true; only one egg, for instance, is usually laid by the sexual female. But this is not always the case, for, according to Buckton, more than one egg is sometimes laid. Such a condition shows that the one egg habit is of much more recent development than in the Eriosomatinae. More- over, the sexual female is very different in structure. She is not the 6 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. deo-enerate, beakless, nonfeeding individual of the Eriosomatiiiae, but feeds upon the leaf like the viviparous forms. Moreover, in external structure these insects resemble certain ones of the Callipterini, to such an extent, indeed, that Davidson has described one of these forms as a Chaitophorus. Certain of the structures met with in these forms resemble those found in the Hormaphidinae and on these structures the tribe might be placed there. The writer believes, however, that the true affinities of the tribe are shown by comparison with the Phyllaphidina. It will be seen at a glance that there is a very close resemblance in all main characters. But the sexual forms are different, although not so strikingly different as would appear at first. . In the Phyllaphidina both winged and wingless ovipara and as a rule mnged males occur. But in some species (quercifoliae) intermediate and apterous males also occur. It is not a very long call, therefore, from the apterous males and females of the Phyl- laphidina to the apterous sexes of the Thelaxini. But the habit of egg laying met with in this last tribe shows that it has been on this course of development longer than has the Phyllaphidina. The group is therefore considered as a tribe which has separated some- what earlier and yet has paralleled in some ways certain characters of the Phyllaphidina. Continuing with the CaUipterus branch we find two somewhat similar lines of development, the one represented by the Callipterina and the other represented by the Chaitophorina. Both are similar in many regards, but are quite different in the armature, particularly of the antennae. The first subtribe separating from that offshoot represented by the Callipterina is the Phyllapliidina. This seems evident from the fact that the oviparous forms of some species are yet alate, a primi- tive condition found very seldom in the Aphididae. The next oft'- shoot resulted in the Callipterina where the oviparous forms are apter- ous, the cornicles of moderate development, and the wing veins usually not reduced. From this offshoot the Saltusaphidina evi- dently arose. This httle subtribe is closely related to the Callip- terina in many ways, but there are some new developments. In the first place the power of leaping has become developed by the enlarging of the femora. Secondly, both the sexual forms have lost their wings, which the male usually retains in the Callipterina. One of the most important points, however, is the fact that in the Sal- tusaphidina the ocular tubercles which represent the retained larval eyes are absent, whereas they are quite conspicuous in the Callip- terina, On this same CaUipterus branch, but somewhat more specialized than the Calhpterina, are two subtribes. These have specialized in opposite directions, the one toward the elimination of the cornicles GENERIC CLASSIFICATION OF APHIDIDAE. 7 and the other toward the development of them. The first subtribe, the Monaphidma, lacks cornicles above. The second of these two, the Drepanaphidina, possesses them in varying degrees. In this last subtribe the males are winged, and the females have developed an extremely long, narrow ovipositor. Coming now to that line represented by Chaitophorus, the Chai- tophorina are found to be the most generalized, corresponding quite closely with the Callipterina. In this subtribe males are winged as a rule, but sometimes in the same species they are intermediate or apterous. Arising from the same branch with the Chaitophorina are two subtribes specialized in different directions, like the sub- tribes of Callipterus. The first, Fullawayina, lacks cornicles entirely, whereas the Pterocommina has developed them in varying degrees, as has the Drepanaphidina. This concludes the subtribes of the Callipterus branch. In connection with these insects the tribe Greenideioi should be considered. The cornicles of the primitive aphids were evidently small, somewhat rounded or conical, and armed with hairs. In the Greenideini the insects have very long cylindric or somewhat swollen cornicles which are thickly covered with prominent hairs. No such well-developed cornicles are met with in any of the other tribes of the family, although they are approached in the Macrosiphina. In this latter subtribe species occasionally occur which show a few short hairs on the cornicles. It seems evident then that the Greenideini separated from the Aphidinae before the hairs of the cornicles disappeared. This was evidently more recent than the development of the tribe Lachnini which possesses a much more primitive cornicle. At about the same time that the ancestors of the CaUipterini separated from the Aphidinae, other forms probably separated and more or less paraUjcled in some w- ays the ancestors of the Macrosiphina, but unlike them carried the hairs of the cornicles. They thus resulted in forms with very long cornicles similar to those of the Macrosiphina but armed with long hairs. In other characters, too, they of course differ, particularly in regard to the cauda. In consideriag the further development of the Aphidinae, a more or less distmct development of the cornicles and antennal tubercles is found. There are thus two types which separate themselves, rep- resented by Aphis and Macrosiphum respectively. These may be considered as leaving the apliid line at about the same time after the development of prominent cornicles. There are, consequently, two, subtribes, the Apliidina and the Macrosiphina. The Cervaphidina represents a group of insects armed with long, somewhat cylindric cornicles, and very prominent spinelike protuberances. The number of antennal segments is somewhat reduced, as is also the whig vena- tion. It seems evident then that this is a subtribe on somewhat the 8 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. same line of development as the Aphidina but developing these spec- ialized spines during the same period in which the wings and antennae have become reduced. Still another subtribe, the Pentalonina, shows a very peculiar wing venation. This is not so much a primitive wing as a more specialized one. It is placed, therefore, as one of the highest subtribes t)f the Aphidini. There remains yet for discussion the tribe Setaphidini. Tliis, it seems evident, belongs with the Aphidmae. In regard to the an- tennsB and the wings it is quite highly specialized but in regard to the cornicles, cauda, and anal plate this statement can not be made. The natural position of this tribe is somewhat doubtful. Its ancestors evidently separated from the aphidian line before the prominent corn- icles of the Aphidina, Macrosiphina, etc., appeared and yet the species are more specialized in many ways than are members of those sub- tribes. It would appear that the lines separated after that of the Greenidemi, for the cornicles are not hairy. Yet this separation must have taken place a considerable time before that of the Aphidina and Macrosiphina. The tribe is placed, therefore, as indicated in the diagram (fig. 1). The subfamilies, other than the Aphiduiae, include the most spec- ialized members of the fanuly. By far the most primitive of these subfamilies is the Mindarinae. This subfamily, as has been indicated, is a remnant from the past, givmg some idea of the ancestors of the Eriosomatinae and the Hormaphidinae. The wing structure is partic- ularly worthy of study. The wing of no other living aphid is like it, but tliis peculiar structure is abundantly met with in fossU forms. The media, it is true, is more reduced than in certain members of the Aphidinae, but this is of very little importance as compared Vvdth the wing's peculiar structure. The form also feeds upon conifers and this is undoubtedly a primitive habit. The cauda and anal plate are imlike those met with either in the Eriosomatinae or the Hormaph- idinae. The sexual forms are interesting. They have become sufficiently specialized toward the Eriosomatinae to have lost the wings, but they retain the beak, at least in most individuals, and feed. The ovaries of the oviparous female also are developed so that a number of eggs are laid. The two remaining subfamilies are the most highly specialized of all aphids. The Eriosomatinae are m many ways more specialized than the Hormaphidinae, but in other ways they are more primitive. The whole Eriosoma line separates at once on the sexual forms. These are small, apterous, and beakless. Throughout their life they take no nourishment, and the ovaries of the oviparous female become atrophied, so that only one develops and of the eggs therein only one GENERIC CLASSIFICATION OF APHIDIDAE. 9 reaches maturity. The most j^rimitive tribe on tliis line is the Eriosomatini. The forms of this tribe are not as a rule distinct gall formers. They possess rather prominent cornicles and have devel- oped special wax glands. They live as a rule upon deciduous trees, the summer forms of many species alternating upon the roots of plants. More specialized than the Eriosomatini are the Pemphigini, which, however, are very similar to the former in many respects. These are distinct and true gall formers on deciduous trees. For part of the year they are usually altogether closed within the gall. Wax secre- tion is common and the cornicles are present, but reduced to mere rings. The Melaphini are closely related to the Pemphigmi and are gall formers like them. These forms, however, have lost entirely the cornicles which are usually stiU retained in the Pemphigini. A somewhat different specialization is met with in the Prociphilini. Here wax secretion has developed at the expense of the cornicles so that these organs are absent, at least in nearly all the forms of the species. Large wax plates have takfen their places. The species are not true gall formers, but live upon foliage which they cause to roU or crumple into a pseudogall. Development along this line is also pres- ent in the next tribe, the Fordini. Here the cornicles are also absent, being replaced by large wax glands, but the species are nearly all root feeders and are usually associated with ants, often living with them in their nests. This tribe may be considered the most specialized of all the Eriosomatinae. The same specialization in the sexual forms has not occured in the Hormaphidmae. They are small and apterous, it is true, but they possess beaks, they feed, and the oviparous female lays more than one egg. In one regard, however, these insects are more specialized. Many of them have developed a pecular aleyrodiform stage, which is quite different from anything occuring elsewhere in the family. Along with this development peculiar wax glands have made their appear- ance so that some of these forms look very much like aleyrodids and are indeed often mistaken for them. The most primitive tribe here is the Oregmini, which, although it possesses many of the other characters met with in these forms, lacks the aleyrodiform stage. These insects possess quite distinct cornicles. Closely related to the Oregmini is the Cerataphidini. These insects likewise possess cornicles and in several ways suggest the Oregmini, but they have developed a distinct aleyrodiform stage and in this regard are much more advanced than the members of that tribe. Lastly, and perhaps most specialized of all, are the Hormaphidini. These insects are curious gall formers, not only on their primary host, but often on their secondary one as well. They lack cornicles and 10 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. have developed ale}Todiform generations and wax secreting structures. In many ways the specialization of these insects is most remarkable. Key to the Subfamilies of the Aphiiudae. 1. Sexual forms small \nih functioning mouth parts absent. Oviparous female with all the egg tubes present or indicated in the embryo but the adult possessing only one tube and maturing one cell so that one egg only is laid. Cornicles much reduced or absent. Wax glands abundantly developed. Wing veins usually reduced. Antennal sensoria prom- inent Eriosomatinae. Sexual forms with functioning mouth parts. Nearly all the ovarian tubes developed in the adult oviparous female 2. 2. Radial sector of forewing inserted mesad of the stigma. Sexes small. Ovi- . parous female laying several eggs Mindarinae. Radial sector not so inserted but arising from the stigma 3. 3. Forms usually gall makers. Wing veins much reduced so that the media is usually simple. Wax glands usual. Antennal sensoria annular. Aley- rodiform stages common. Sexes wingless as a rule and small. .Hormaphidinae. Forms not usually gall makers. Wing veins often not reduced. Wax glands not abundant. Antennal sensoria oval or subcircular. Aleyrodi- form stages rare. Cornicles often little reduced. Winged males common. Aphidinae. Subfamily I, APHIDINAE. The subfamily Aphidinae contains many of the most primitive insects in the family. Indeed, with the exception of the Mindarinae the subfamily may be considered as by far the most primitive. The oviparous female, in all the tribes, develops the ovaries in a normal way and lays several eggs. An exception to this, however, is the Thelaxini, but here two or more eggs are sometimes laid. The males may be either alate, apterous, or mtermediate, and in many species which possess the migratory instinct they are often produced on quite a different food plant from the oviparous form. The stem mothers are in practically all cases apterous, but the remaining gener- ations throughout the year may or may not be winged. In many species a larger percentage of winged forms occurs in certain generations and a larger percentage of apterous forms in others. In some species, however, this does not appear to be the case. In certain of the Callipterini practically all of the viviparous forms other than stem mothers are winged. The insects are mainly foliage feeders, but they also attack the stems and roots. They occur both upon woody plants and herbs. Their feeding may have little apparent effect upon the host or it may cause distortions or pseudogalls. Some species are particularly in- jurious to their hosts and when these are economic plants cause much loss. Great variation is met with amongst the members of the subfamily. The antennse are rather long and slender and as a rule are armed with subcircular sensoria. In most of the forms the sixth segment GENEEIC CLASSIFICATION OF APHIDIDAE. 11 possesses an elongate narrow- unguis, which in some of the CaUipterini and Aphidini is remarkably developed. In the more primitive groups, however, this is short and thumb-like. The head of th^ apterous form differs much from that seen m the Eriosomatinae, in that true compound eyes are present and often very prominent, and the small larval eyes are seen as ocular tubercles. It is noteworthy that in the Eriosomatinae the alate forms possess distinct compound eyes but the apterous forms have lost them. The wings are in general quite similar throughout the family in regard to the venation. In color, shape, and location of the veins there is often considerable difference. Moreover, there are a few genera amongst the different tribes which show abnormal wing form, of which genera Microparsus is a good example. In the typical forms of this subfamily the media of the fore wing is twice branched, but it is very commonly branched only once and it is rarely simple. The cornicles show remarkable variation. In some forms of the CaUipterini they are short and slightly swollen at the base, in the Lachnini they are low broad cones, whereas in the Greenideini they are cylindrical and sometimes longer than the body. Between these extremes every gradation occurs. The cornicles may be straight oi* they may be swollen to a greater or less degree. Practically all forms eject a colored wax from these organs when disturbed. The Cauda shows almost as much variation as the cornicles, some- times being short and rounded, in other cases elongate, spatulate, or conical, and in others distinctly knobbed. Variation also is met with in the anal plate, though this usually is rounded. In the CaUip- terini, however, it is often bilobed. The tribes of the subfamily may be separated as foUows: Key to the Tribes of the Aphidinae. 1. Cornicles situated on broad flat cones 2. Cornicles truncate, or more or less elongate 3. 2. Cornicles and antennas hairy. Antennae with the unguis short and thick. . Lachnini. Cornicles and antennae not hairy. Antennae with the unguis long and slender Setaphidini. 3. Cornicles clothed vrith. long hairs Greenideini. Cornicles never with long hairs 4. 4. Thorax of alate form with the lobes not prominently developed; oviparous form small, often laying one egg. Large wax plates present Thelaxini. Thorax of alate form with the lobes prominently develoj^ed; oviparous female laying several eggs. Large wax plates usually absent 5. 5. Cornicles truncate or elongate; when elongate the cauda knobbed, and the anal plate bilobed, or the antennae prominently hairy Callipterini. Cornicles not truncate, usually elongate. Cauda never knobbed. Anten- nae with only a few spinelike hairs Aphidini. 12 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Tribe LACHNINI. The tribe Lachnini is the most primitive of all living aphids, with the exception of the Mindarinae. The genus Mindarus shows in its wing structure characters more primitive than any of the Lachnini, but in other characters such as those of the beak, cornicles, cauda, sensory structures, etc., the Lachnini are very primitive insects. In examining the fossil wings it is to be noted that the radial sector is situated back of the stigma. In practically all living aphids, with the exception of Mindarus, this vein has migrated toward the tip of the wing. In primitive forms the stigma is long and narrow, whereas in most living forms it has become more or less compact. In the subtribe Lachnina the radial sector has become a very short, straight vein almost at the tip of the wing. This shows that the Lachnina are evidently more advanced than the Pterochlorina in which the radial sector is somewhat curved and situated near the middle of the stigma. The subtribe Eulachnina is considerably specialized, as indicated by the eyes, the shape of the body, and the cornicles. It is, however, as closely related to the Lachnina as are any of the other tribes, as will be seen from the formation of the wing. The Anoecina in the typical genus shows a wing with a short blocky stigma, a condition quite different from that seen in the Lachnina, and the radial sector is here curved. (In Nippolachnus, however, the stigma is still long and straight.) Moreover, the sexual forms are more specialized, being apterous in both cases. Anoecia, therefore, is somewhat removed from Lachnus. The genus Trama is considerably specialized, in that it lacks cornicles in the apterous form. It is, however, related to Lachnus through Neotrama with small cornicles, and Protrama with large hairy cornicles. The rostrum in the Lachnini is in many species five-segmented, a primitive character most marked in this group. The freshly emerged wing of a lachnid shows that Mj, M2, and M3+4 are the veins repre- sented when the media is twice branched, and that in some species no vein is formed about Mj. The cubitus and first and second anal arc present in the forewing. As in other Aphididae, however, no vein forms about the second anal. The radial sector is in Lachnus a short, straight trachea and a prominent vein forms about it. The stigma, as in all members of the family, is formed b}^ radiusj. In the hindwing both media and cubitus are present and form distinct veins. The antennae of the Lachnini are six-segmented with a short unguis. They are usually armed with oval or subcircular sensoria and prominent hairs. In fact, the entire body of the insect is hairy. The cornicles are characteristic. They are situated on distinct cones which are constricted before the somewhat flanged opening which is not situated over the center of the cone. The cones are arm- GENERIC CLASSIFICATION OF APHIDIDAE. 13 ed with hairs. Some specialized forms have small cornicles or none at all. Wax-secreting structures, but no distinct gland areas, are present in this tribe and a coating of fine wax is often found over the entire insect, including the appendages. This is true of the oviparous forms, as well as of the viviparous ones. The Cauda and anal plate are here rounded, never developed into elongate structures as in some of the other tribes of the subfamily. The sexual forms are nearly as unspecialized as the viviparous ones. Both sexes possess a distinct rostrum and take food. The males in the typical subtribes are winged. The females are apterous, but the ovaries are developed and several eggs are laid by each individual. Key to the Subtribes op the Lachnini. 1. Radial sector of fore wings curved and of moderate length 2. Radial sector of fore wings short and straight, situated near the tip of the wing 4. 2. Hind tarsi extremely elongate, head divided, wing venation usually faint. Tramina. Hind tarsi normal 3. 3. Stigma short and thick, sexes both apterous Anoecina. Stigma elongate, males often winged Pterochlorina. 4. Form elongate and very narrow; antennae with bristles, cornicles not hairy; eyes without ocular tubercles Eulachnina. Form notelongate; cornicles on hairy cones; eyes with ocular tubercles. .Lachnin A. Subtribe ANOECINA. The subtribe Anoecina is suggestive of the Tramina, but none of the forms are as specialized as some of the genera of that subtribe. The typical genus is quite distinctive in the short rounded stigma and in the sexual forms.. The genus Nippolachnus, however, has a stigma quite Lachnus-like in appearance. Only two genera are known at present. Key to the Genera op thp Anoecina. 1. Head not divided; eyes with prominent ocular tubercles; stigma of wing short and rounded Anoecia. 2. Head divided; eyes without ocular tubercles; stigma long and straight. Nippolachnus. Genus ANOECIA Koch. Plate 1, A-F, I. 1857. A noecia Koch, Die Pflanzenlaiise Aphiden, p. 275. Characters. — Head not divided, front somewhat rounded. Eyes prominent but not distinctly set off from the head. Antennae of six segments, armed with subcircular or oval or elongate sensoria and covered with hairs. Fore wings with the media once branched. Stigma short and thick. Hind wings with both media and cubitus present. Cornicles situated on broad hairy conea. Cauda and anal plate somewhat rounded. Spring forms free, living in colonies; summer forms often subterranean. Sexes small and apterous, possessing beaks and feeding. Oviparous female laying one or more than one egg. Type (monotypical), Aphis corni Fab. 14 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Genus NIPPOLACHNUS Matsumura. Plate I, G, H, J, K. 1917. Mppolachnus Matsumura, Journ. Coll. Agr. Tohokvi Univ., v. 7, pt. 6, p. 382. This genus is closely related to Anoecia Koch and yet it retains many Lachnus characters which are not present in Anoecia. It is separated from that genus by several important points. The head is distinctly divided as it is not in Anoecia. The stigma is long and Lachnus-like and the eyes have not the small prominent ocular tubercles of Anoecia, but are rounded on their posterior margins. Characters.— Antennee of six segments, armed with prominent protruding sensoria and many hairs; head di\'ided; eyes without prominent ocular tubercles. Wings with the media once branched, the stigma rather long and narrow. Cornicles on large broad cones entirely covered with hairs. Spring forms free, migrating in summer to alternate hosts. Sexes small, males winged. Type (fixed by Matsumura, 1917), Nippolachnus piri Mats. Subtribe EULACHNINA. The subtribe Eulachnina is related to the Lachnina quite closely, as can be seen by the wing structure where the radial sector is a straight, short vein extending across the tip of the wing. The media also is faintly indicated. In the other branch of the tribe in which the Anoecina and Tramina are found, the radial sector is curved as it is in Aphidina, etc. Members of the Eulachnina may, however, be separated at once from the Lachnina on the elongate, narrow shape of the body, the abruptly rounded cauda, the character of the cornicles, and the absence of ocular tubercles upon the eyes. The subtribe is evidently quite specialized as compared to the Lach- nina. Characters. — Eyes large and set off fi'om the head; ocular tubercles not CAident; antennae slender, armed with bristles or spines, not slender hairs. Cornicles shallow, not on distinct hairy cones. Cauda abruptly rounded. Body very elongate and slender, scarcely wider than the head. Key to the Genera of the Eulachnina. li Antennse of five segments, armed with minute bristles Essigella. Antennae of six segments 2. 2. Media once branched, antennse with long stout spines Eulachnus. Media twice branched Todolachnus. Genus ESSIGELLA Del Guercio. Plate I, S-Y. 1909. Essigella Del Guercio, Rivista Patol. Vcgetale, n. s., v. 3, p. 329. The genus EssigeEa is quite similar to Eulachnus with the excep- tion of the antennae. Characters. — Head with large outstanding eyes, very much broader than long. Antenna; of five segments, imbricated, armed only with a few minute bristles. Fore wings Avith the media faintly indicated, once branched, hind wings with both media and cubitus present. Cornicles chitinized rings situated close to the body, no hairy cones present. Cauda rounded. Body elongate and narrow. Type (monotypical), Lachnus califarnicus Essig, GENERIC CLASSIFICATION OF APHIDIDAE. 15 Genus EULACHNUS Del Guercio. Plate I, L-R. 71853. Cinaria Curtis, British Entomology, v. 12, section 576. 1909. Eulachnus Del Guorcio, Ri\ista Patol. Vegetale, n. s., v. 3, p. 329. 1915. Protolachnus Theobald, Bull. Ent. Res., v. 6, p. 145. Del Guercio erected the genus Eulachnus without setting a type, but Wilson * has indicated agilis Kalt. as the type. Apparently, therefore, the genus must be based upon that species. Theobald's genus was based on his tuherculostemmata, a species in which the char- acters are the same. Cinaria was erected with pini L. as type, but this was questioned. Characters. — Head divided, eyes rather large and outstanding; antennae of six seg- ments, armed with long stout bristles. Fore wings with the media faintly indicated and once branched ; hind wings with both media and cubitus present but faint; corniclea minute rings, not situated on hairy cones. Cauda abruptly rounded. Body elongate and narrow. Type (fixed by Wilson, 1911), Lachnus agilis Kalt. Genua TODOLACHNUS Matsumura. 1917. Todolachnus Matsumura, Jour. Coll. Agr. Tohoku Univ., v, 7, pt. 6, p. 381. The writer has been unable to study the type of this genus, but from the description given it seems to represent a genus belonging here and having a twice-branched media. The description of the cornicles as "wart-like, not broader at base" would indicate its affinities here, also the words ' ' body long, nearly parallel on the lat- eral sides." Type (fixed by Matsumura, 1917), Todolachnus abietis Mats. Subtribe LACHNINA. The members of the subtribe Lachnina may be separated from those of other subtribes with the exception of the Eulachnina by the charaQter of the venation. The radial sector here has almost reached the tip" of the wing and become a short straight vem. The genera may be separated as follows: Key to the Genera of the Lachnina. 1. Media of fore wings twice branched Dilachnus. Media not twice branched 2. 2. Media once branched 3. Media simple Unilachnus. 3. Labium lance-like Lachnus. Labium ol^tuse Schizolachnus. Genus LACHNUS Burmeister. 1835. Lachnus Burmeister, Handbuch der Ent., v. 2, pt. 1, p. 91. 1909. LachnieUa Del Guercio, Redia, v. 5, p. 286. The genus Lachnus Burmeister was erected with the following included species: lapidarius Fab., /agri Linn., quercus Uiim., fasciatus Burm., and punctatus Burm. Of these species the following were removed as types of other genera: fagi, 1857, PliyllapJiis, and quercus, 1870, StomapMs. 1 Wilson, H. r. Notes on the synonymy of the genera included in the tribe Lachnini. In Ann. Ent. Soc. Amer. v. 4, p. 51-54, 1911. 16 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. The following type fixations have ])een made for Lachnus: 1840. ApMs roboris Linnaeus, Westwood. 1863. Lachnus pinicola Kaltenbach, Passerini. 1908. ApMs nudus De Geer, Mordwilko. 1910. Lachnus punctatus Burmeistcr , Wilson. 1911. Lachnus fasciatus Burmeister, Wilson. Now the first three fixations are invalid, since the species were not included in the original genus. The first valid fixation, there- fore, is that of Wilson, 1910, when he set punctatus as type. This fixation, according to present rules, can not be changed in 1911 because it is an unfortunate fixation, but punctatus must remain the type of the genus 'Lachnus. The question is now purely zoo- logical. At present punctatus is unknown and, therefore, the genus Lachnus must remain unknown until punctatus is discovered. This is the situation, if the rules are followed, and the well-known genus name will be lost to us. At the suggestion of numerous aphid w^orkers we are holding fasciatus as the type of Lachnus and the Commission wUl be asked to suspend the rules in this case on account of the long usage of the name Lachnus. In 1909 Del Guercio erected the genus LachnieUa without setting a type but in 1911 WUson interpreted this genus as Lachnus with fasciatus Burm. as ty^^e. FoUowiug this the writer definitely desig- nated this species as type. Therefore, LachnieUa will become a syno- nym of Lachnus. Characters. — Eyes large, with distinct ocular tubercles present. Antennse of six segments and with rather prominent hairs. Cornicles on somewhat shallow hairy- cones. Fore wings with the radial sector short and straight; stigma elongate; media once branched. Labium lance-like. Type (by suspension of rules), /nsciatus Burm. Genus DILACHNUS Baker. Plate II, A-C. 1919. Wilsonia Baker, Can. Ent., v. 51, p. 212. 1919. Dilachnus Baker, Can. Ent., v. 51, p. 253. Characters. — Eyes with distinct ocular tubercles. Antennae of six segments and armed with slender hairs and circular sensoria. Cornicles on rather broad hairy cones. Fore wings with radial sector straight, media twice branched ; hind wings with both media and cubitus present. Type (fixed by Baker, 1919), LachnieUa gracilis Wlsn. Genus SCHIZOLACHNUS Mordwllko. Plate II, D. 1908. Schizolachnus Mordwilko, Ann. Mus. Zool. I'Acad. Imp. des Sci., St. Petersbourg, v. 13, p. 375. The genus Schizolachnus was erected by Mordwilko with tomen- tosus De Geer (j^ineti Fab.) as type. Characters. — Eyes large and with ocular tubercles present. Antennte of six seg- ments and with rather prominent hairs. Cornicles on somewhat shallow hairy cones. Fore wings with the radial sector straight; stigma elongate; media once branched. Labium obtuse. Type (fixed by Mordwilko, 1908), Aphis toinentosus De Geer. GENERIC CLASSIFICATION OF APHIDIDAE. 17 Genus UNILACHNUS Wilson. Plate II, 'E-G. l'J19. Unilachnus Wilson, Ent. Xews, v. 30, p. 5. The genus Unilachnus Wilson appears to be a connecting link between this subtribe and the Eulachnina. In some respects the genus is very suggestive of that subtribe. The form is elongate and the cornicles are reduced. They are armed, however, with hairs and are not so reduced as in the Eulachnina. The ocular tubercles here are very rudimentary, almost absent, and in this regard, too, the genus suggests the Eulachnina, but it seems to belong in this subtribe. Characters. — Form elongate; cornicles somewhat reduced; ocular tubercles small; media of fore wings simple. Antennee of six segments moderately armed. Tj'pe, Lachnus parvus Wilson. Subtribe PTEROCHLORINA. The genera l)elonging in the subtribe Pterochlorina are in some ways more primitive than those of the Lachnina, but in other respects some of them are more specialized. The radial sector of the fore wings is still curved and in some genera quite elongate. This is much nearer the early type of wing than is the wing of the Lachnina where the radial sector is short and has migrated almost to the tip of the wing. Of course, the distinct curving of this vein found in some of the genera is an advance on the slightly curved elongate vein usually met with in the fossils, but to our mind the location and character of this vein are much more primitive than in the Lachnina. The stigmal area and the sexual forms appear considerably more primitive than in the Anoecina, the specialization of which has been in a different direction from that of the Lachnina. The male of Stomaphis is, however, an exception. Characters. — Head often divided; antenme of six segments, armed with hairs and subcircular sensoria. Fore wings V\-ith radial sector somewhat curved and not close to the tip of the wing. Cornicles on broad hairy cones. ^lales usually winged. Key to the Genera of the Pterochlorina. 1. Stigma extending along costal margin almost to the tij) of the wing..Longistigma. Stigma not so extending 2. 2. Beak extremely long, very much longer than body; antenme covered with very fine, short hairs Stomaphis. Beak normal in length; antennaj with rather stout hairs, often quite long Pterochlorus. Genus LONGISTIGMA Wilson. Plate II, H-L. 1909. Longistigma AVilson, Can. Ent., v. 41, p. 385. 1909. Davisia Del Guercio, Redia, v. 5, p. 185. The genus Longistigma Wilson can be distinguished at once by the shape of the stigma which is drawn out at the tip to an acute point which extends almost to the tip of the wing. The type species is 141613°— 20— Bull. 826 2 18 ' BULLETIN 826^ V. S. DEPARTMENT OF AGRICULTURE. voiy large, one of the largest aphids. Del Guercio set no type for his subgenus which was published shortly after Wilson's. Wilson placed this species as a synonym of caryae Harris which he made the type of Longistigma. Characters. — Size large. Head somewhat divided. Eyes large, with distinct ocular tubercles. Antennae of six segments, armed with subcircular sensoria and prominent hairs. Fore wings with media twice branched, radial sector not a great way from the tip of wing and stigma extending around almost to tip; hind wings with both media and cubitus present. Cornicles on broad, shallow, hairy cones. Cauda and anal plate somewhat rounded. OviparoTis female apterous. Males winged. Tj'pe (monotypical). Aphis caryae Harris. Genus PTEROCHLORUS Rondani. Plate II, S-X. 1848. Pterochlorus Rondani, "Familia Hcmipterorum AphidiniB " in Nuovi Annali delle Scienze Naturali, p. 35. 1855. Dryobius Koch, Die Pflanzenliuise Aphiden, p. 225. 1908. Tuberolachnus Mordwilko, Ann. Miis. Zool. de I'Acad. Imp. des Bel. St. Petersbourg, v. 13, p. 374. 1909. Dryaphis Del Guercio, Redia, v. 5, p. 262. 1913. ScMzodryobius Van der Goot, Tidj. voor Ent., v. 56, p. 130. 1917. Pterochlorides Archangelsky, Turkestan Ent. Stn. Rept. Tashkent. 1918. TuberodryoUus Das, Mem. Ind. Mus., v. 6, p. 259. The generic name Cinaria was used for Aphis pini L ?, and Aphis rohoris L. with A. pini as type. In the writer's opinion Cinaria can not be used now with rohoris as type. Rohoris was used as the type of Dryobius, therefore this name is clear. Dryaphis was used with Pterochlorus as a synonym but the name as a generic name was really first used by Del Guercio in 1909. Tuberolachnus was erected in 1908 with viminalis Boyer as type but the difference in the abdominal tubercle is not, in the writer's opinion, sufficient for a distinction. Van der Goot's genus is plainly a synonym. The other two generic names listed were used with Lachnus persicae Choi, as type. This species has several abdominal dorsal tubercles but for the same reason as Tuberolachnus is held to be a synonym. Characters. — Head somewhat rounded. Antennas of six segments, armed with subcircular sensoria and prominent hairs. Fore wings with radial sector distinctly curved, inserted some distance from tip; media twice branched. Hind wings with both media and cubitas present. Wings often banded or mottled, cornicles on hairy cones. Cauda and anal plate rounded. Abdomen sometimes with dorsal tubercles. Tj^e (fixed by Rondani, 1848), Aphis rohoris Fab. (=roboris L.). Genus STOMAPHIS Walker. Plate II, M-R. 1870. Stomaphis Walker, The Zoologist, v. 28, p. 2000. 1881. Rhynchodes Altum, Forst Zool., v. 3, p. 356. Of the five species originally in the genus Lachnus the species quercus Limi. was removed by Walker as the type of his genus Stoma- phis. Little confusion has arisen in regard to this species. Altum's genus was erected for Rhynchodes longirostris. GENERIC CLASSIFICATION OF APHIDIDAE. 19 Characters. — Head slightly rounded; eyes large with distinct ocular tubercles. Antennae of six segments armed with large sub circular sensoria and thickly covered with fine hairs. Cornicles situated on very broad, shallow, hairy cones. Cauda sub- conical, slightly rounded . Anal plate rounded. Fore wings with radial sector rather long and somewhat curved. Media twice branched; hind wings with both media and cubitus present, quite widely separated. Beak very long, much longer than body. Males wingless and with rudimentary mouth parts. Type (fixed by Walker, 1870), Aphis quercus Linn. Subtribe TRAMINA. The subtribe Tramina is composed of insects quite specialized in nature, subterranean and often associated with ants. The typical genus Trama is the most specialized of all and is in some char- acters suggestive of the Fordina. Its relations with the other Lachnini, however, are shown clearly by the other genera. The genus Trama, as described by Del Guercio (Redia, v. 5) possesses small cornicles. Specimens of troglodytes, however, received from Schouteden, Mordwilko, and others lack cornicles entirely. Some other forms possess them either as very small cones or as large Lachnus-like structures. These latter are evidently the most primi- tive and to these is given the name Protrama. The insects with small cornicles are grouped under the name Neotrama. The genera may be separated as follows : Key to the Genera of the Tramina. 1. Apterous form entirely without cornicles and with rudimentary eyes Trama. Ai3terous form with cornicles 2. 2. Apterous form with large, broad, Lachnus-like cornicles and large distinct compound eyes Protrama. Apterous form with small cone-like cornicles armed with a few hairs and more or less rudimentary eyes Neotrama. Genus PROTRAMA, n. gen. riate III, r-T. Head divided, front straight, eyes i>rominent and set off from the head. Antennse of six segments armed with hairs and small subcircular protruding sensoria. Cornicles situated on broad, low, hairy cones. Cauda and anal plate rounded. Hind tarsi extremely elongate. Wing venation faint; fore wings with the media twice branched ; hind wings with both media and cubitus present. Type, Trama radids Kalt. Genus TRAMA, Heyden. Plate III, N. 1837. Trama Heyden, Mus. Senkb., v. 2, p. 293. Head divided but not prominently so, front straight. Apterous form with the eyes reduced to a few facets. Antennse of six segments; cornicles absent; cauda subcorneal, rounded. Anal plate rounded. Entire insect minutely hairy. Type (mono typical), Trama troglodytes Heyden. 20 BULLETIN 826, U. S, DEPAETMENT OF AGRICULTURE. Genus NEOTRAMA, n. gen. Tlate III, M, O. Head somewhat flat. In the ai^terous form the eyes reduced ; antennae of six segments and the cornicles on very small cones with a few scattered hairs. Cauda sul) conical, rounded. Anal plate rounded. Entire body covered with fine hairs. Hind tarsi greatly elongate. Type, Tramn troglodytes Del Guercio {^Neotrama delguercioi Baker). Tribe THELAXINI. It has been the custom of most writers to place the Thelaxini (Vacuini) in the Eriosomatinae, often possibly because of the fact that only one egg is laid by the oviparous female. But the female is quite different in structure from the beakless females of the Eriosomatinae and the other forms are very different indeed. With the Hormaphidinae there are more resemblances, the most striking of which is the structure of the thorax. The mesothorax indicates very faintly the lobes so prominent in most forms. The presence of distinct cornicles, however, is very different from the forms in the Hormaphidinae lacking these although possessing a some- what similar thorax. The sensory structures, too, are widely different, being similar to those found in the Phyllaphidina. Indeed, the antennae are very like those of that subtribe. The oviparous forms of the Phyllaphidina, however, lay several eggs and may be either winged or apterous and the males, though sometimes apterous, are usually winged. The venation of the Thelaxini is more reduced than in the Phyllaphidina. Taking all of these facts into consideration it seems evident that the Thelaxini should be placed in the Aphidmae and somewhat related to the Phyllaphidina, a subtribe which belongs in the Callipterini. It is evident, however, that the Thelaxini must stand somewhat apart; it is placed, therefore, as a tribe of the subfamily Aphidinae next to the Callij)terini. In this tribe the specialization of the ovipara has advanced beyond that of the Callipterini in that only one egg is laid, but according to Buckton several eggs may be laid and the distinct beak is evidence of relationship. Characters. — Cornicles present as chitinized rings on shallow hairy cones. An- tennaj somewhat setose, with oval or subcircular sensoria. Cauda somewhat semi- circular or distinctly knobbed. Body usually armed with hairs or stout spines. Sexual forms small and apterous, possessing beaks; oviparous female as a rule laying only one egg. Forms li\'ing free upon the foliage. The genera may be separated by the following key: Key to the Genera op the Thelaxini. Cauda distinctly knobbed Thelaxes. Cauda not knobbed but somewhat semicircular Glyphina* GENERIC CLASSIFICATION OF APHIDIDAB. 21 Genus GLYPHINA Koch. riatc III, G-L 1857. Glypkina Koch, Die Pflanzenlause Aphiden, p. 359. 1911. Travaresiella Del Guercio, Redia, v. 7, p. 299. ( haracters. — Com^'cles present as somewhat elevated rings. Antermx 5-seg- inented, minutely setose, armed with a few stout hairs and somewhat subcircular fiensoria. Fore wings with the media once branched; hind wings with only the media present. Cauda not knobbed, eomewhat rounded, anal plate rounded. Body covered with hairs. Forms living upon the foliage ot plants. Type (monotypical), Glyphma heticlae Kalt. Genus THELAXES Westw. I'late III, A-F. Vacuna of authors, not Hcydcn. 1840. Thelaxes Westw., Int. Mod. Class. Ins. Synopsis, v. 2, p. 118. In 1837, Heyden erected his genus Vaouna based on coccinea Heyden. He definitely stated that he thought Phylloxera Boyer was the same genus. Kaltenbach stated that coccinea is a Phylloxera and so considered dryopMla as type of Vaouna, as this species was in- cluded in the genus by Heyden. On the authority of Schouteden and other European workers coccinea is now considered a Phylloxera and another type, dryopliilc/,, can not be set for the genus in order to apply Vacuna to the genus as now understood. Vaouna with coccinea as type will become a synonym of Phylloxera and another name will be necessary to apply to the genus having dryopMla as type. The next name used appears to be Thelaxes Westwood. Characters .—Gorniclea present as chitinized rings on broad Ioav cones. Antenna? of the stem mother 5-segmented. Alate form vdth. 3-segmented antennas, sensoria oval or subcircular. Fore wings M-ith the media once branched, hind wings with the cubitus lacking. Cauda distinctly knobbed, anal plate rounded. Sexual forms small and apterous, possessing distinct beaks and feeding; oviparous female producing normally but one egg. Type (fixed by Westwood, 1840), Thelaxes quercicola Westw.(=Aphis dryopMla Schr.) Tribe CALLIPTERINI. The tribe Calliptermi is composed of forms which live upon the foliage of plants. The species in many of the sub tribes have developed peculiar habits. Some forms are almost solitary whereas others live in colonies. Some have developed the power of leaping, while others are sedentary. The sexual forms do not vary greatly from the viviparous forms. In nearly all of the sub tribes the males are winged, though in the Saltusaphidina they are apterous. In the other tribes intermediate males may occur in the same species with alate males. The oviparous females are nearly always apterous, although in the Phyllaphidina alate ovipara may occur. Both sexes feed and the ovaries of the oviparous female are developed so that several eggs are. laid. 22 BULLETIN 820; U. S. DEPARTMENT OF AGRICULTURE. The wing veins are not reduced in most species. In some^ however, the radial sector of the fore wings is very faint or entirely absent. This condition is met with in several genera of the Callipterina. That it is this vein which is lacking is indicated by the trachea of the freshly emerged A\ing. Here the media is represented by M^, M2, and M3+4. The cubitus and first anal are distinct trachese, whereas the second anal is faintly indicated. In the hindwing besides the radial sector three oblique tracheas are present; these are the media, cubitus, and first anal. Only the media and cubitus are represented in the vena- tion. Considerable variation is met with in the cornicles of this tribe but they are never long and prominent as in the Aphidini. The usual form is the truncate one represented in Myzocallis, Chaitophorus, etc. Very often the cornicles are sculptured. In some cases they are reduced to small cup-shaped structures and in others they are represented by mere rings. The antennae, as a rule, are long and slender and armed w4th few sensoria. These sensoria are usually small, subcircular or oval. In rare cases they are somewhat elongate. The Cauda in this tribe is as a rule knobbed and the anal plate bilobed. In some cases, however, the cauda and anal plate are both rounded. In the Saltusaphidina the anal plate is divided and the cauda remains distinctly knobbed. Wax secretion is present to a limited extent in this tribe. It is most developed in the Phyllaphidina. Here there are large lateral ab- dominal wax plates in all of the forms and the insects present a wool- like appeaj-ance on the foliage. In the genus Euceraphis wax secretion is found to a limited extent. In one species, mucidus Fitch, it is, however, abundant and the insects of this species often seem to float in the air, a peculiar appearance cdmmon also in the Erioso- matinae. In the Saltusaphidina also distinct wax plates occur, particularly in the oviparous forms. These are arranged along the abdominal segments. The habit of leaping is common in the Saltusaphidina as the name implies. Here the muscles of the femora are greatly enlarged for this purpose. Many of the other members of the tribe approach this con- dition, especially in the genus Monellia. Others, although they do not distinctly leap, drop so suddenly when disturbed that they almost ap- pear to leap from the foliage. Our common Symydobius on the birch is difiicult to collect on account of such a habit and other forms of Callipterina are very similar in action. Certain species in this tribe are closely attended by ants in return for the honeydew excreted. Some species are protected by these Hymenoptera by means of sheds or roofs built over colonies on the leaves or twigs. These sheds are f omid quite commonly upon the leaves of the oaks protecting the spe- GENEEIC CLASSIFICATION OF APHIDIDAE. 23 cios described as Symydohius alhasiphus by Davis and here placed as Neosymydobius. The internal structure of insects of this tribe appears not to differ markedly f roni the structure in other groups. Witlaczil, however, has reported that in certain members of this tribe the intestine forms a closed loop almost similar to that found in the Chermidae. As a rule, in this tribe, the various forms met with in the subfamily occur. In the genus Monellia, however, in some species at least, ap- terous viviparous forms seldom occur, nearly all the viviparous forms being alate. The sub tribes may be separated by the following key: Key to the Subtribes of the Callipterini. 1. Eyes with ocular tubercles present, head not elongate 2. Eyes without ocular tubercles present, head often elongate Saltus aphidina. 2. Antenn£e armed with rather long, prominent hairs 3. Antennae usually only with minute, sometimes stout bristles 5. 3. Cornicles absent • Fullawayina. Cornicles present 4. 4. Cornicles cylindrical or vasiform Pterocommina. Cornicles truncate, enlarged at base Chaitophorina. 5. Cornicles absent above Monaphidina. Cornicles present, position as usual G. 6. Cornicles reduced to mere rings; large lateral abdominal wax i:)lates present Phyllaphidina. Cornicles usually not reduced to mere rings; no large abdominal wax plates present 7. 7. Cornicles variable, often long and somewhat swollen; oviparous female with an elongate ovipositor Drepan aphidina. Cornicles never long; always short and truncate; oviparous fem?le not always with an elongate o\'ipositor Callipterina. Subtribe PHYLLAPHIDINA. The subtribe Phyllaphidina is erected for the species related to the genus Phyllaphis. Many of the characters show these species as quite closely related to the Callipterina, while in other ways they very strongly suggest the Thelaxini, as indicated under the discussion of the tribe. C/jaraciers.— Cornicles present; antennae of six segments, minutely setose, sensoria elongate or subcircular; cauda knobbed or rounded, anal plate often bilobed, wax glands present. Forms living free or in pseudogalls. Sexual forms often alate, some- times, however, apterous or intermediate, showing that the apterous condition has developed but recently ; oviparous female producing several eggs. Key to the Genera of the Phyllaphidini. 1. Anal plate deeply cleft and U-shaped Shivaphis. Anal plate entire or somewhat bilobed, not deeply cleft 2. 2. Cauda rounded, anal plate entire 3. Cauda knobbed, anal plate somewhat bilobed Phyllaphis. 3. Oviparous females with annular sensoria Neophyllaphis. Oviparous females with small transverse sensoria Tamalia. 24 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Genus NEOPHYLLAPHIS Takahashi. 1920. Ncophyllaphis Takahashi, Can. Eat., v.!}2,Tp. 20. Characters. — Cornicles slightly elevated. Antennae of six segments, with narrow, transverse sensoria. Fore wings \vith media twice branched, hind wings with both media and cubitus present. Cauda and anal plate rounded, cauda sometimes slightly constricted. Oviparous females winged and possessing annular sensoria. Forms wax secreting and living free on the plants. Type (monotypical), Neophyllaphis podocarpi Takahashi. Genus PHYLLAPHIS Koch. Plate IV, FF, GG. 1S57. Phyllaphis Koch, Die rflanzenliiuse Aphiden, p. 248. The well-known genus Phyllaphis Koch is represented by/a^iL., but as indicated under Tamalia has been made to include, species of somewhat different structure. Characters. — Cornicles present as chitinized rings which are very slightly elevated on low conical bases. Antennae of six segments, long and slender, minutely setose, sen- soria narrowly oval. Fore mngs with the media twice branched ; hind wings with both media and cubitus faintly indicated. Cauda knoblied, anal plate slightly divided. Forms living upon the foliage, sometimes producing a curling of the leaves. Males usually winged ; o^'iparous form apterous, producing several eggs. Type (monotypical). Aphis Jr. gi L. Genus TAMALIA n. gen. Plate IV, HH, II. The genus Tamalia Baker is erected for Pemphigus coiceni Ckll., a species which has since been placed in Phyllaphis, It is, however, quite distinct from the type of the genus and undoubtedly represents a now genus. Mr. Theodore Pergande received this species, coiveni, from California and made some notes on the material, thinking, how- ever, that it was a new species. He gave it a provisional name and the new generic name here used. This name, as far as our knowledge goes, never was published. Among the large number of new genera conceived by Pergande this is one of the few valid ones and his man- uscript name, therefore, is used here. Characters. — Cornicles present as mere flangeson low, broad, conical bases. Antennae of six segments minutely setose and with narrow sensoria. Fore wings with media once branched, hind wings with both media and cubitus present. Cauda and anal plate both rounded. Abdominal wax plates present. Forms living in pseudogalls. Sexes sometimes l)oth winged, oviparous form pro- ducing several eggs. Type, Pemphigus coiveni Ckll. Genus SmVAPHIS Das. 191S. Shiujphis Das, Mem. Ind. Mils., vol. (>, p. 245. The genus Shivaphis was erected with celti Das as type, and cclti is evidently the species rodescribed as Chwrnayliis celticolens by Essig.^ 1 EssiG, E. 0., and Kuwana, S. I. Some Japanese Aphididaj. In Proc. Cal. Acad. Sci., v. S, no. 3, p. 95, 1918. GETiTERIC CLASSIFTCATTOlSr OF APHIDIDAE. 25 The writer lias studied a series of specimens of this species taken in 1907 on Celtis sinensis. There seems little doubt that the genus is related to Phyllaphis. The dorsal wax glands are of much the same structure as those found in fagi L. The deeply cleft anal plate, however, at once separates the two. The cauda whix^h is almost cylindrical in some specimens is quite distinctly knobbed in others. Characters. — Cornicles present as mere rings. Head without prominent antennal tubercles. Antennae of six segments, sen soria elliptical. Fore wings with the media twice branched, hind -wings with both media and cubitus present. Cauda cylindrical or knobbed. Anal plate deeply dhdded. Males winged. O^dparous females apter- ous. Rows of wax glands present. Type (monotypical), Shivnphis celti Das. Subtribe CALLIPTERINA. Tlie subtribe Callipterina is a somewhat large and interesting one. Some of the species are very large and more or less solitary, others are small and live in colonies. In some genera apterous viviparous forms, with the exception of the stem mother, seldom occur, while in other genera they are as a rule, present. The males are in most cases alate and the oviparous forms apterous. The species of nearly all the genera live upon the leaves of plants. They do not affect these greatly, as a rule, although when abundant the insects often seriously interfere with the proper development of the trees attacked. Many of the insects are armed upon their bodies with prominent spines or tubercles. Characters. — Cornicles present, truncate in form. Antenna with setae or spines, of six segments, and armed with subciroular or in a few cases somewhat elongate sen- soria. Wings often clouded, mottled, or banded. Cauda as a rule knobbed, anal plate usually more or less indented or bilobed. Body often armed with capitate spines or tubercles. Key to the Genera of the Callipterina. 1. Cauda distinctly knobbed, anal plate usually bilobed or sometim.es deeply divided 3. Cauda not distinctly knobbed, anal plate entire or almost so 2. 2. Antenn;ie minutely setose, sensorium at base of unguis oval or somewhat rounded; oviparous female with secondary sensoria Symydobius. Antenna? not minutely setose, sensorium at base of unguis long and narrow, oviparous female without secondary sensoria Euccrapbis. 3. Anal plate deeply divided with a U-shaped cleft so that the lobes appear as distinct; cauda knobbed Therioaphis. Anal plate bilobed, not deeply divided; cauda very markedly knobbed 4. ' 4. Antennte and often the cornicles with prominent hairs Callipterus. Cornicles and antennae without such hairs 5. 5. Cornicles much reduced ; Mdngs sometimes horizontal in repose Monellia. Cornicles truncate, fairly well developed; wings not horizontal in repose. . . 6. 26 BULLETIN 82<}^ U. S. DEPARTMENT OF AGRICULTURE. G. More or less distinct antennal tubercles present; oviparous female with secondary sensoria Calaphia. No distinct antennal tubercles present ; oviparous female without secondary sensoria 7. 7. Anal plate slightly indented, sometimes almost entire; no apterous vivi- parous forms developed Chromaphis. Anal plate distinctly bilobed; apterous viAdparous forms common Myzocallis. Genus CALAPHIS Walsh. Plate IV, S, U. 1863. CalapMs Walsh, Proc. Ent. Soc. Phila., v. 1, p. 301. 1913. Siphonocallis Del Guercio, Redia, v. 9, p. 293. 1913. CaHiptcrinclla Van der Goot, Tijdschr. voor Ent., v. 56, p. 118. Walsh erected his genus CalapMs for his hetulella, a species which lacks the radial sector in the wing, and on this character he based his genus. Del Guercio based his Siphonocallis on hetulaecolens Fitch, distinguishing it from his conception of the genus Callipterus, which conception, however, was not according to type. A study of hetulaecolens shows that in all respects, with the exception of the radial sector, this species is similar to hetulella. Many specimens of hetulae- colens lack the radial sector and in most cases it is only faintly indi- cated at best. These two species, therefore, are probably congeneric. The genus Callipterinella was based on hetularius Kalt., and this species proves to be very similar to hetulaecolens. It is true that the frontal tubercles are not prominent in tliis species as they are in the type of Calaphis. There seems no doubt, however, that all of these three species are closely related. This relation is shown in part by the sexual forms. The oviparous females all possess sensoria on the antemiaB and are very similar in other body characters. This pres- ence of sensoria in the oviparous form, while not important in some groups, separates quite distinctly this small group of species from those of the Myzocallis type. It seems evident that the relations of hetu- larius are with hetulaecolens and hetulella. Callipterinella, therefore, is also a synonym. Characters. — Cornicles present, distinct, truncate. Antennae of six segments, armed with oval sensoria and placed on more or lesa distinct tubercles. Fore wings with the media twice branched, the radial sector either absent or faintly indicated, sometimes, however, complete; hind wings with both media and cubitus present. Cauda distinctly knobbed, anal plate bilobed, body with prominent hairs. Forms living more or less solitary upon the foliage, sexes not markedly different from the other forms; oAdparous female producing several eggs and possessing sensoria upon the antennaj. Type (monotypical), Calaphis hetulella Walsh. fr GENERIC CLASSIFICATION OF APHIDIDAE. 27 Genus CALUPTERUS Koch. Plate IV, I, J. 1855. CaUipterus Koch, Die Pflanzenlaiise Aphiden, p. 208. 1870. Callaphis Walker, The Zoologist, v. 5, p. 2000. 1881. Ptychodes "Buckton, Mon. British Apliids, v. 3, p. 39. 1904. Panaphis Kirkaldy, The Entomologist, v. 37, p. 279. 1917. Nippocallis Matsumura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 3(35. In erecting the genus CaUipterus, Koch included a number of species among which w^as juglandis Kalt. In 1860, Passerini set juglandis as the type of CaUipterus and erected his MyzocaUis for species similar to coryli Goetz. Most WTiters overlooked Passerini's work and considered the genus CaUipterus in the light of MyzocaUis. This is the conception com- monly held by many to-day. The appUcation of CaUipterus, how- ever, must be restricted to species essentially like juglandis. All the other generic names listed as synonyms were, with one exception, used with this same species juglandis as type and therefore require little comment. The genus Nippocallis was erected with Icuricola Mats, as type. Specimens of this species studied by the writer^ are not in good con- dition for the observation of the anal plate. All the characters visible, however, indicate that this species is a CaUipterus. Characters. — Cornicles present, truncate in form, rather prominent and often armed with long hairs. Antennse of six segments, armed with stout hairs, sensoria usually- oval. Fore wings with the media twice branched, hind wings with both cubitus and media present. Veins usually bordered. Radial sector often faintly indicated. Cauda not distinctly knobbed in all cases. Anal plate bilobed ; body usually covered with prominent hairs. Forms living free upon the foliage. Sexual forms not differing markedly from the other forms, oviparous female producing several eggs. Type (fixed by Passerini, 1860), Aphis juglandis Frisch. Genus CHROMAPHIS Walker. Plate IV, O, P. 1870. Chromaphis Walker, The Zoologist, v. 5, p. 2001. The genus Chromaphis was erected with juglandicola Kalt. as type. It 'is related quite closely to MoneUia. Characters. — Cornicles moderate in size, somewhat flanged. Antennse of seven seg- ments armed with oval sensoria. Fore wings with the media twice branched, hind wings with media and cubitus present, wings not held horizontally in repose. Cauda knobbed, anal plate slightly indented. Sexual forms somewhat similar to the vivip- arous ones. Males usually winged. Oviparous female with the ovaries developed normally, laying numerous eggs.^ Forms living free upon the foliage usually all summer; \dviparous generations winged. Type (monotypicalj, Aphis juglandicola Kalt. > Davidson believes as many as 30 may be produced by one female. 28 BULLETIN 820, U. S. DEPARTMENT OF AGRICULTURE. Genus THERIOAPHIS Walker. Plate IV, K, I.. 1870. Thcrioaphis Walker, The Zoologist, p. 1999. 190.5. KalUslajMs Kirkaldy, Can. Ent., v. 37, p. 417. 1906. Eucal'iptcrus Schouteden, Ann. Ent. Soc. BoIe;., v. 50, p. 31. i 1915. Neocallipterus Van der Goot, Beitrage zur Kennt. der lloll. Blattlause, p. 320. The genus Therioaphis Walker was erected with ononidis Kalt. as type, and ononidis has been shown by Theobald to ])e the common "yellow clover aphis," trifolii of Monell. This species has a deeply cleft anal plate quite different from that of Myzocallis. Eucallipterus was erected with tiliae L. as type, a species with quite similar struc- ture. Eucallipterus, therefore, will become a synonym. BetuUcola Kalt. has been used as type by Kirkaldy and Van der Goot. Accord- ing to Das, Van der Goot considers this congeneric with trifolii. Therefore Therioaphis is the name that must be used. Characters. — Cornicles truncate, rather constricted mesad of apex. Antenna cf six segments without prominent hairs and armed vdth subcircular sensoria. Fore wings with media twice Ijranched ; hind \rings with both media and cubitus present. Wings often variously marked. Prothorax rather elongate; cauda knol)bed. Anal plate deeply bifid so that two long, narrow lobes are formed. Body often with prominent hairs. Type (monotypical), Jphu ononidis K<. Genus EUCERAPHIS Walker. I'latc IV, Q, R. ' 1S70. Euceraphis Walkor, Tho Zoo'.o-ist, p. 2001. 1908. CalHptcroides Mordwiiko, Ann. Mus. Zool. I'Acad. Imp. des f'ci. St. Petersbourg, v. 13, p. 377. 1913. Callipteroidcs Van der Goo:, Tijd. voor. Ent., v. 56, p. 151. When Walker erected his genus Euceraphis with hetulae L. as type he had in mind evidently the same species as that described by Koch under the same specific name, and thus separated species of this type. Mordwilko in 1900 erected the genus Callipteroides with nigritarsus Hey den as type. Specimens of tliis species received from Mordwilko show that the species he had was the hetulae of Koch or at least a species very close to it. Tliis would then make Callipte- roides a synonym of Euceraphis. In 1913 Van der Goot used the name CaUipteroides with hetulae Koch as type and his placing, there- fore, should ])e under Euceraphis. Characters. — Cornicles present, truncate. Antennae of six segments, long and slender, armed with rather narrow sensoria usually near the base of segment III, the unguis of segment VI usually not much longer than the base, sensorium at the base of unguis long, oval, a nd fringed ; more or less distinct frontal tubercles present. Fore wings with the media twice branched, hind wings with both media and cubitus present. Cauda usually knobbed and rather large. Anal plate usually entire. Abdomen of the alate form often with distinct wax-producing glands. Forms very large and usually solitary in habit, sexes similar to the other forms; oviparous female producing se\'eral eggs. Type (fixed by Walker, 1S70), Aphis hetulae (L.) Walker {=Callipterus betulae Koch). 1 There is considerable evidence for keeping this genus distinct. GENERIC CLASSIFICATION OF APHIDIDAE. 29 Genus MONELLIA Oestlund. Plate IV, M, N. 18S7. MonclUa Oestlund, Gsol. and Nat. Hist. Survey Minn., Bui. no. 4, p. 41. The genus Monellia Oesthmd was erected for caryeUus Fitch and only one species was included in the genus at the time. Several other species, however, were made synonyms of caryeUus which are quite different from that species and fall into other genera. Characters. — Cornicles present as mere rings; antennae slender, of six segments; sen- soria oval or subcircular; head broad for its length; prothorax prominently separated from the mesothorax. Fore wings with the media twice branched ; hind wings with both media and cubitus present. Wings often held flat upon the back, cauda knobbed, anal plate Vjilol^ed. Forms li\4ng solitary upon the leaves, sometimes having the power of leaping. Apterous forms rare. Sexes feeding; oviparous female laying several eggs. Type (monotypical), Aphis caryella Fitch.. Genus MYZOCALLIS Pass. Plate IV, G, H. 1860 1860 1894 1894 1913 1915 1917 1917 Myzocallis Passerini, Gil Afldi, p. 28. Pterocallis Passerini, Gli Afidi, p. 28. Subcallipterus Mordwilko, Varshava Univ. Izvlestiia, v. 8, no. 58, p. 53. Tuberculatus Jlordwilko, Varshava Univ. Izvlestiia, v. 8, no. 58, p. 60. CalUpterus Van der Goot, Tijd. voor Ent., v. 56, p. 116. Tubcrculoides Van der Goot, Beitrage zur Kennt. der Holl. Blattlause, p. 313. AcanthocaUis Matsumura, Jour. Coll. Apr. Tohoku Univ., v. 7, pt. 6, p. 368. Takecallis Matsumura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 373. When Passerini erected the genus Myzocallis in 1860 he placed coryli Goetz as type. Later in the same work he erected his genus Pterocallis with alni Pass, as type. This species proves to be very similar indeed to coryli, so similar in most of the characters that the writer believes the genus Pterocallis to be a synonym of Myzocallis. The type of the genus, coryli Goetz, was also placed as the type of Callipterus by Van der Goot, in 1913, in spite of the fact that another type for that genus had been set in 1860. Callipterus Van der Goot (1913) is therefore a synonym of Myzocallis. The species alni Pass, is universally considered the same species as aZm Fab., and this species was made the type of the genus Subcallipterus by Mordwilko in 1894. Subcallipterus Mordwilko, 1894, is therefore a synonym of Myzocallis for the same reason as is Pterocallis Pass., 1860. The species querceus Kalt. was made the type of the Tuberculatus by Mordwilko in 1894, but this species seems too'closely related to coryli. Tuberculatus, therefore, becomes a synonym. The species quercus Kalt. was made the type of Tuberculoides by Van der Goot, 1915, and this is quite typically a Myzocallis. A number of genera have been erected by Matsumura which are so very little different from the type species that they are listed here as synonyms. This author follows the idea of proportions of the antennal segments as generic characters. 30 BULLETIN 826; U. S. DEPARTMENT OF AGRICULTURE. Characters.— Cormcles truncate mthout a very distinct neck; antenna^ of six segments armed with a few minute bristles and oval or subcircular sensoria. Fore wi°ng8 with the media twice branched ; hind wings with both media and cubitus present. Cauda knobbed; anal plate bilobed, not di\Tided; body usually with stout hairs. T>T)e (fixed by Passerini, 1860), AjMs coryli Goetz. Genus SYMYDOBIUS Mordwilko. Plate IV, DD, EE. 1894. Symydohius Mordwilko, Varshava UniTersitetskiia Iz^4estiia, v. 8, no. 58, p. 65. 1917. Yezocallis Matsumura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 369. Symdobius of later authors. The o-cnus Symydobius Mord., which was erected for oUongus Heyden, has often been spelled Symdobius by subsequent writers. This is probably due to the erection of the genus in a Russian publi- cation which is available to few workers, at least in this country. Specimens of this species studied were collected by Mordwilko at Petrograd and Warsaw and by Schouteden at Brussels. The species which passes under the name of oUongus in America is quite distinct, as has been pointed out by the writer. Characters. — Cornicles present, truncate or with an evident neck and on a broad low base. Antennse of six segments armed with numerous delicate hairs, sensoria somewhat oval or subcircular; sensorium at the base of the unguis not long and nar- row, with a fringe but without a prominent one, cauda semicircular, anal plate similar in shape, sometimes slightly indented. Fore wings with the media twice forked; hind wings mth both media and cubitus present, , somewhat separated at the base. Type (monotypical), Aphis ohlonga Heyden. Subtribe SALTUSAPHIDINA. The subtribe Saltusaphidina is separated from the other related ones principally on the nature of the head. The most important character, possibly, is the structure of the eyes, m which the ocular tubercles appear to be wholly lackmg. Characters. — Forms living usually in damp places upon the foliage of sedges and grasses, narrow elongate bodies, eyes with ocular tubercles lacking, legs often modi- fied for leaping. Oviparous forms apterous, somewhat similar to the vi\'iparous forms, producing several eggs. Key to the Gener.\ of the Saltusaphidina. Head considerably elongate, cornicles cup shaped, legs modified for leap- ing Saltusaphis. Head not much elongated, cornicles mere rings, legs not modified for leap- ing Thripsaphia. Genus THRIPSAPHIS Gillette. Plate TV, X. 1917. ThripsapMs Gillette, Can. Ent., v. 49, p. 193. The genus Thripsaphis was separated from Saltusaphis for halli Gillette, which the present writer had included in that genus, and certain other similar species. GENERIC CLASSIFICATION OF APHIDIDAE. 31 Characters. — Cornicles j^resent as slightly elevated rings; antennae of six segments armed with subcircular sensoria. Eyes ■without ocular tubercles. Fore wings with the media twice branched, hind wings with the cubitus sometimes absent, cauda knobbed, anal plate divided, body with spinelike hairs. Forms living free upon grasses and sedges in moist localities. Sexes apterous, oviparous female producing several eggs. Type (fixed by Gillette, 1917), Brachycolus balli Gill. Genus SALTUSAPHIS Theobald. Plate IV, V, W. 1915. Saltusaphis Theobald, Bull. Ent. Research, v. 6, pt. 2, p^ 138. Characters. — Cornicles present, cup-shaped or truncate; antennae of five segments, minutely setose; sensoria small and subcircular. Head elongate, ocular tubercles absent. Fore wings with the media twice branched; hind wings with the cubitus usually absent; cauda knobbed; anal plate divided, caudal extremity of the abdomen sometimes bilobed; body covered with spines which are often modified into different shapes. Forms living more or less solitary upon the leaves of grasses or sedges in marshy regions. Sexual forms apterous, oviparous female laj'ing several eggs. Type Tmonotypical), Saltxisaphis scirpus Theo. Subtribe DREPANOSIPHINA. The subtribe Drepanosiphina is evidently related to the callip- teriiie branch of the tribe rather than the chaitophorine one. It has specialized in the opposite direction from the Monaphidina in that the cornicles are more or less prominently developed. It would appear to bear the same relation to this branch of the tribe as does the Pterocommina to the chaitophorine one. There is considerable variation in the development of the cornicles, even within certain of the genera. Some have very large cornicles, others have small ones. They all appear, however, to have the same general structure. Characters. — Cornicles present, vaiying greatly in development from very small to very large. Cauda somewhat knobbed, anal plate slightly indented. Oviparous female with a long drawn out ovipositor. Key to the Genera of the Drepanosiphina. 1. Cornicles extremely long and somewhat swollen in the middle .Drepanosiphum. Corniclesnotextremelylongand larger at the base 2. 2. Cornicles very small, truncate Neosymydobius. Cornicles large with a swollen region at the base Drepanaphis. Genus DREPANAPmS Del Guercio. Plate IV, JJ-LL. 1909. Drepanaphis Del Guercio, Rlvista di Patologia Vegetale, n. s., v. -1, no. 4, p. 49-50. 1909. Phymatosiphum Davis, Annals Ent. Soc. America, v. 2, p. 196. The two generic names given were both used with the same type species, and therefore no discussion in regard to the use of the names is necessary. Characters. — Cornicles large but not of the same shajje as those of Drepanosiphum, being rather narrow toward the distal extremity and swollen at the base. Antennae of six segments armed with subcircular or oval sensoria and a few scattered hairs. Fore wings with the media t\rice branched, hind wings with both media and cubitus present. Cauda knobbed, anal plate somewhat indented. Forms living more or less 32 BULLETIN 826;, U. S, DEPARTMENT OF AGRICULTURE. solitary upon the foliage of trees. 'Males winged. Cniparourj female with a distinct elongated o\dpositor and producing several eggs. Type (monotypical\ r>rfpano!fiphum acerifoJii Thos. Genus DREPANOSIPHUM Koch. Plate IV, MM. 1S55. Drcpanosiphum Koch, Die Pflanzenlause Aphiden, p. 201. 1885. Type fixation, Lichtenstein, Monographie des Aphidicns, p. 175. Cornicles very long, quite distinctly swollen in the middle or subcylindric. An- tennge of six segments with short, scattered hairs and oval or subcircular sensoria. Fore ^vings ^\'ith the media t\vice branched. Hind mngs with both media and cubitus present. Cauda knobbed, anal plate slightly indented. Forms lining upon the foliage of plants, males usually winged. Oviparous female Avith a distinct elongated o\'ipositor. Type ('fixed by T.ichtenstein, 1885"), Aphis flatanoides Schr. Genus NEOSYMYDOBIUS, n. gen. The genus Neosymyclo])iiis is erected for species similar to that described as S ymydohius alhasiplius Davis. It is evident that this species is not a Symydobius. Members of that genus are very large and differ in several wa3's. We place the present genus here with considerable doubt. Characters. — Cornicles small, truncate, Callipterus-like. Antenme of six segments which are armed with a few rather stout hairs. Fore wings with the media twice branched, hind ^vings mth both media and cubitus present. Cauda knobbed, anal plate slightly indented. Forms li\-ing in colonies upon the foliage of trees. Body, particularly of the apterous forms, coa ered with rather stout spine-like hairs. Ovipar- ous female with a long o^'ipositor and depositing several eggs. Males usually winged. Type, Symydohius alhasiphus Da\'ie. Subtribe MONAPHIDINA. The subtribe Monaphidiiia, erected for the genus Monaphis, is very similar in most respects to the Callipterina, but lacks the cornicle dorsally. It seems evident that it is a specialization from insects of the Callipterus type in much the same way that Fullawaya is related to those of the Chaitophorus type. One genus only is known at present. This differs more from the Callipterina than does Fullawaya from the Chaitophorina. Genus MONAPmS Walker. 1870. Monaphis Walker, The Zoologist, p. 2001. 1894. Bradyaphis Mordwiiko, Varshava Universitetskiia Izvicstiia, v. 8, p. 59. Few remarks on the synonymy of the genus Monaphis Walker are necessary as the same species was used as type in both cases men- tioned. The genus is a very remarkable one, being peculiar in many ways. Characters. — Corn^les faint; antennae of six segments, without distinct haire, sensoria small and circular; cauda somewhat rounded but with an acute point or pro- jection; anal plate similar; fore wings with the media twice branched ; hind wings with both media and cubitus present. Type (fixed by Walker, 1870), Apnis c.ntennctn Kalt. GENERIC CLASSIFICATION OF APHIDIDAE. 33 Subtribe CHAITOPHORINA. The subtribe Chaitophoriiia is composed of apliids which are similar ill many ways to the Calhpterina. They differ in that they are always armed with long hairs which quite prominently cover the antennae as well as the other parts of the body. Some of the species perhaps are given more to living in colonies than are the Callipterina, but this habit varies in that subtribe as well. Key to the Genera' of the Chaitopiiorina. 1 . Cauda quite distinctly knobbed 2. Cauda not knobbed but rounded 3. 2. Antennaj of fiA'e segments Sipha. Antennae of six segments Chaitophorus. 3. Body elongate ; small dimorphic forms developed Periphyllus. Body not elongate; no dimorphic forms developed 4. 4. Antennae of five segments Atheroides. Antennae of six segments 5 5. Anal plate entire or slightly indented Neothomasia. Anal plate divided into two quite separate parts Patchia. Genus ATHEROIDES Haliday. 1S39. Atheroides Haliday, Ann. and Mag. Nat. Hist., v. 2, p. 189. More careful collecting of marsh-inhabiting species may show that this genus is a specialization from the Chaitophorma as is the Saltu- saphidina from the Callipterma, for most specimens falling here seem to lack ocular tubercles. Characters. — Antennae of five segments armed with stuut spines. Fore wings with media twice branched; hind wings with both media and cubitus present. Cornicles reduced to mere rings. Cauda broadly rounded. Form elongate and flat. Entire insect prominently spined. Species living on sedges and grasses. Type (set by Kirkaldy, 1906), Atheroides serrulatus Haliday. Genus CHAITOPHORUS Koch. Plate IV, CC. 1854. Chaitophorus Koch, Die Pflanzenlause Aphiden, p. 1. 1870. Tranaphis Walto'r, The Zoologist, p. 1999. 1870. A rctaphis Walker, The Zoologist, p. 2000. 1912. Eichochaitophorus Essig, Pom. Coll. Joum. Ent., v. 4, p. 721. 1912. Micrella Essig, Pom. Coll. Journ. Ent., v. 4, p. 716. 1856. T3T)e fixation, Gerstaecker, Bericht for 18-54, p. 1('>2. In 1854 Koch erected the genus Chaitophorus with several species. Aphis populi was made the type by Gerstaecker in 1856. The same species was used by Walker as type of his genus Arctaphis and there- fore this name will become a synonym. The types of both of Essig's genera show that these vary little from populi. The genus Tranaphis was erected with salicivorus Walker as type and this species is similar in general characters to populi. Therefore, Tranaphis will become a synonym. 141613°— 20— Bull. 826 — —3 34 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. Ill 1S60 Passerini used aceris L. as the type of Chaitophorus and this placing has often been followed, but that of Gerstaecker has priority. Characters.— Cornicles present, truncate, rather prominent. Antennae of six seg- ments, armed ^vith subcircular sensoria and rather prominent hairs. Fore mngs mth the media normally twice branched, hind wings ^vith both media and cubitus piesent. Cauda distinctly knobbed. Anal plate entire, sometimes somewhat indented. Sex- ual forms not differing markedly from the \i\-iparous ones. Males winged, as a rule, but sometimes intermediate or apterous. O^-iparous females apterous with the ovaries normally developed and producing several eggs. Both sexes feeding. Forms living usually upon the leaves of trees; no dimorphic forms developed. Type (fixed by Gerstaecker, 1856), ApJris popuH L. Genus PATCHIA, n. gen. Characters. — Cornicles truncate; antennae of six segments, hairy and with circular sensoria. Fore wings with the media twice branched, the radial sector absent or faintlv indicated; hind wings with both media and cubitus present. Cauda rounded or slightly conical; anal plate divided into two separate parts. T\^e, Patchia rirginiann Baker. Patchia virginiana, n. sp. Alate viviparous female. — Antennae as follows: III, 0.48 mm., with an even row of about 12 subcircular sensoria; lY, 0.288 mm.; V, 0.24 mm.; VI (0.16-0.192 mm.). Color brown with a large black patch on dorsum of abdomen and with lateral patches of same color. Wings ^vith the radial sector absent and the veins heavily bordered. Apterous form almost solid velvety black. Both forms secreting wax. Found on the bark of chestnut at East Falls Church, Va. The type is in the U. S. National Museum (Cat. No. 23063^. Genus PERIPHYLLUS Van der Hoeven. Plate IV, A A, BB. 1852. Phillophorus Thornton, Proc. Ent. Soc. London, n. s., v. 2, p. 78. 1858. Chclymorpha Clark, The Microscope. 1863. Pcriphyllus Van dcr Hoeven, Tijd. voor. Ent., v. 6, p. 7. 1913. Chaitophorinclla Van der Goot, Tijd. voor Ent., v. 56, p. 150. 1917. Arakawana Matsumura, Journ. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 375. In 1852 Thornton used the name Phillophorus with his testudinatus as type. This name had, however, been used in 1840. Koch erected tlie genus Chaitophorus in 1854 and included therein a number of species. In 1856 Gerstaecker set Apkis populi L. as the type of Chaitophorus and therefore prevented the use of the name for spe- cies such as testudinatus unless aU of Koch's species are included. In 1858 Clark used the name Chelymorpha with the specific name 2)liyllo2)liora. The species he discussed is the testudinatus of Thorn- ton. The generic name Chelymorpha, however, was used as early as 1834 and, therefore, is not available. In 1863 Van der Hoeven employed the generic term Periphyllus with his species testudo as type. This name is a synonym of testudinatus Thornton, and the generic name seems to be the first one available. In 1913 Van der Goot employed the generic name Chaitophorinella with testudinatus as type, and this name, therefore, will become a synonym of Periphyllus. GENERIC CLASSIFICATION OF APHIDIDAE. 35 Characters. — Cornicles present, truncate iu form, often sculptured. Antennse of six segments (with the exception of the dimorph) armed with oval sensoria and promi- nent hairs. Fore wings with the media twice branched; hind wings with both media and cubitus present. Cauda and anal plate rounded. Forms living upon the foliage of trees. Sexes not strikingly different from the other forms, possessing beaks and feeding. Males winged, oviparous females with the ovaries normally developed, thus laying several eggs. Small lamellate or hairy dimorphic forms produced in summer. Type (monotypical), Pcriphyllus tcstudo Van der Hoeven (=^testudinatus Thorn- ton). Genus NEOTHOMASIA, n. n. riate IV, Y, Z. 1910. Thomasia Wilson, Can. Ent., v. 42, p. 386. Wilson erected the genus Thomasia with populicola Thos. as type, and his description appeared in December, 1910. The same name had, however, been used for a genus of Diptera, the description of which appeared m September, 1910. A new name, Neothomasia, therefore, is necessary for Wilson's genus. Characters. — Cornicles present; antennae of six segments armed with subcircular sensoria and prominent hairs. Fore wings with the media twice branched, liind wings with both media and cubitus present. Cauda and anal plate both rounded. Forms living in colonies upon the leaves or bark of trees; no dimorpliic forms pro- duced; sexual forms not markedly different from the vi\dparous ones. Oviparous females laying several eggs. Type (monotypical), Chaitojjhorus populicola Thos. Genus SIPHA Pass. 1860. Sipha Passeriiii, Gli Afldi, p. 29. This genus and Atheroides are distinct from the other genera iu the sub tribe by possessing five-segmented antennse instead of six- segmented ones. The genus has not been much confused excepting by Thomas's placing of ruhifolii. For a time some workers in this country were led to conceive of the genus as indicated by that spe- cies which in reality belongs m the Aphidini. Characters. — Cornicles present, truncate, short, almost mere tings. Antemase of five segments armed with large circular sensoria. Body form flat, entire insect cov- ered with rather long stout hairs. Fore wings with the media twice branched, hind wings with both media and cubitus present. Cauda knobbed, anal plate rounded. Forms living upon the leaves of grasses usually in moist localities, sometimes even submerged, the water appearing to affect them little. Type (fixed by Passerini, 1S60), Aphis glyceriae Kalt. Subtribe PTEROCOMMINA. The subtribe Pterocommina is composed of bark-feeding insects, some of which retain quite primitive characters. It is the writer's opinion, however, that they are, as a group, more specialized than the Chaitophorina, but closely related. This is indicated by the development of the cornicles met with in the sj^ecies. Like the 36 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. Drepanosipliina, this development varies to a great extent in the different species. Only two genera occur in the tribe. They may be separated as follows : Key to the Genera of the Pterocommina. Cornicles cylindrical Pterocommci. Cornicles somewhat swollen Melanoxantherium. Genus PTEROCOMMA Buckton. Date IV, PP. 1857. Cladohius Koch, Die rflanzenlausc Aphidcn, p. 251. 1860. Aphioidcs Passerini, Gli Afidi, p. 28. 1879. Ptcrocomma Buckton, Monog. Br. Aphides, v. 2, p. 142. 1905. Arislaphis Kirkaldy, Can. Ent., v. 37, p. 416. In 1857 Koch erected the genus Cladobius with fopuleus Kalt. as type. This name, however, had been used previously. So Passerini in 1860 employed the name Aphioides. This name had also been used. Kirkaldy, therefore, gave the new name Aristaphis in 1905. In 1879, however, Buckton described the genus Pterocomma with a very similar species as type. Characters. — Cornicles present, rather short and cylindrical. Antennte of six segments armed with prominent hairs and subcircular sensoria. Fore wings with the media twice branched; hind wings with both media and cubitus present. Cauda and anal plate rounded. Type (monotypical), Pterocomma pilosa Buckt. Genus MELANOXANTHERIUM Schouteden. Plate IV, NN, GO. 1879. ^^danoxanthus Buckton, Monog. Br. Aphides, v. 2, p. 21. 1901. Melanoxanlherium Schouteden, Ann. Ent. Soc. Belg., v. 45, p. 113. In 1879 Buckton described the genus Melanoxanthus with solids L. as type, but, as this name was preoccupied, Schouteden suggested the name Melanoxantherium. Characters. — Cornicles present, variable in size, but usually more or less swollen. Antennae of six segments armed thickly with hairs and possessing oval or sul)circular sensoria. Fore wings with the media twice branched, hind wings with both media and cubitus present. Cauda and anal plate rounded. Forms living in colonies usually on the bark of trees, males usually winged. Ovi- parous females laying several eggs. Type (monotypical), Aphis salicis L.* Subtribe FULLAWAYINA. The subtribe FuUawayina is related somewhat closely to the Chaitophorina from which it is a specialization, as is evidenced by the reduction and total lack of the cornicles. It is very suggestive of Monaphis, but evidently arose from quite a different line of develop- ment, following Chaitophorus rather than the Callipterus group. Only one genus is represented. > The writer is forced to change Ms view that Pterocomma pilosa is closely related to populifoliae Fitch. This was based on Pergande's published statement of his examination of the type. Later notes on the type indicated that it resembles populca. GENERIC CLASSIFICATION OF APHIDIDAE. 37 Genus FULLAWAYA Essi;. 1012. Fullawaya Essig, Pomona Coll. Journ. Ent., v. 4, p. 716. The genus Fullawaya seems to hear somewhat the same relation to Chaitophorus that Monaphis does to Myzocallis. Were it not for the hairy condition of the antennae and the character of the cauda and anal plate, Fullawaya might he placed as a synonym of Monaphis, but it is evidently unrelated. Characters. — Cornicles absent; antennae of six segments, armed with ratlicr long bristle-like hairs, sensoria small and circular, cauda rounded and armed with long curved hairs. Fore wings with the media twice branched, hind wings with Iwth media and cubitus present. Forms li^dng upon the roots of plants. Type (monotypical), Fullawaya saliciradicis Essig. Tribe GREENIDEINI. The trihe Greenideini was first separated by Wilson under the name Trichosiphina. The insects falling here show a most remarkable development of the cornicles. These are sometimes as long as the entire body in the alate forms. In the apterous individuals they are usually swollen. In both they are very thickly covered with long hairs, a condition not met with in any of the other Aphidinae with long cornicles. It is true that some species of Macrosiphum and occa- sionally species of the other genera show here and there minute hairs on the cornicles, but they do not approach in any way members of this tribe in cornicle armature. The development of the cauda in this group is also remarkable. Characters. — Cornicles present and remarkably developed into cylindrical or slightly swollen tubes often as long as the body and thickly covered with long hairs. Antennae of five or six segments armed with oval or subcircular sensoria. Males winged. Forms living free upon the foliage. Key to Gexera of Greexideixi. 1. Antennae of five segments Eutrichosiphum. Antennae of six segments 2. 2. Fore wings with the media twice branched, hind wings with both media and cubitus present Greenidea. Fore wings with media once branched, hind wings with neither media nor cubitus present • Greenideoida. Genus GREENIDEA Schouteden. Plate V, F-K. 1905. Greenidea Schouteden, Spol. Zejlan, v. 2, p. 181. 1906. Trichosiphum Pergande, Ent. News, v. 17, p. 208. In 1905, Schouteden erected a genus for the Si phono pTiora artocarpi of Westwood and redescribed the species, giving details lacking in Westwood's paper. Pergande erected his genus Trichosiphum, making his anonae the type. The characters Pergande used to sep- arate his genus were in reality those of Greenidea. 38 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. ij Characters. — Coraicles extremely long and hairy; antennte of six segments armed ■with oval or subcircular sensoria and distinct hairs. Fore wings with media twice branched; hind wings with both media and cul)itus present. Sexes winged. Type ('monotypicaD, Siphonopkora artorarpi Westw. Genus GREENIDEOIDA Van der Goat. Plate V, L-P. 1900. Greenidea Wilson, not Schouteden, Ann. Ent. Soc. Amer., v. 3, p. 317. 1916. Greenidcoida Van der Goot, Zur Kenntniss der Blattliiuse Java's, p. HO. In discussing the genera of the Trichosiphini in 1910, Wilson based his descriptions and key on specimens in the collection of the Bureau of Entomology. Material in that collection determined as artocarpi by Pergande proves not to be that species for it does not agree with the descriptions given either by Westwood or Schouteden. Wilson, therefore, used the two generic terms Trichosiphum and Greenidea. In reality the species listed as artocarpi was undescribed at that time and the species of Trichosiphum presented all of the characters of Greenidea. Since that time Van der Goot has erected the genus Greenideoida for such species as that understood by Wilson to be artocarin. Characters.— Cornicles present, very long, subcylindrical, and armed with long hairs. Antennee of six segments armed with oval or subcircular sensoria. Fore wings with the media once branched; hind wings reduced in size and lacking both the media and cubitus. Cauda and anal plate rounded. T>T)e (fixed by Van der Goot, 1916), Greenideoida elongata V. d. Goot. Genus EUTRICHOSIPHUM Essig & Kuwana. Plate V, A-E. 1918. Eutrichosiphum Essig & Kuwana, Proc. Cal. Acad. Sci., v. 8, no. 3, p. 97. The genus Eutrichosiphum Essig and Kuwana was erected for the s])ecies pasaniae Okj . Characters. — Similar in general characters to Greenidea. Antennte of five segments^ armed with long hairs and somewhat oval sensoria. Cornicles very long, subcylindrical and covered with hairs. Fore wings with the media twice branched, hind wings with both media and cubitus present. Cauda and anal plate rounded. Type (monotypical), Trichosiphum pasaniae Okj. Tribe SETAPHIDINI. The correct position of the tribe Setaphidini is somewhat difficult to ascertain. In some ways it closely resembles the Aphidini, and in others suggests the Lachnini. It has Aphis-like antennae with, however, a reduced number of segments. The venation of the wings is also reduced. On the other hand it possesses cornicles situated on low flat cones somewhat like those of the Anoecina or Lachnina. But these cones are devoid of hail's. GENERIC CLASSIFICATION OF APHIDIDAE. 39 It would iippcuir that this tribe separated from the aphid line after the cornicles had lost their armature and before their development as indicated in the Aphidina and Macrosiphina began. These organs then remained somewhat primitive whereas reduction took place in the antennae and wings. One genus only is represented. Genus SETAPHIS Van der Goot. Plate V, Q-X. 1916. SetapTiis Van der Goot, Zur Konntniss der Blattlause Java's, p. 153. Characters. — Comiclea present as rings, situated on low broad cones. Antennae of five segments armed with small circular sensoria. Fore wings with the media once branched, hind wings reduced. Cauda and anal plate rounded. Body with two prominent caudal fingerlike projections. Type (fixed by Van der Goot, 1916), Setaphis luteus V. d. Goot. Tribe APHIDINI. The tribe Aphidini is by far the largest tribe of the living Aphididae. Many of the most common species in the family as well as many of the most injurious ones belong here, and it is these forms which correspond to the popular conception of the family. Besides being abundant they are varied, and a large number of genera is therefore found in this tribe. Specialization has taken place in a number of directions, but particularly in the development of cornicles, cauda, etc. The wings have become somewhat reduced in certain genera, but as a rule little reduction in these organs has occurred, the venation in most cases being as complete as in even the most primitive forms of the family. The antennae have developed an elongate filamentous process to the distal segment, which in the Lachnini is represented by a very short thumblike projection. Wax secretion is found scarcely at all apart from that produced and secreted by the cornicles. The head shows certain peculiar develop- ments in some of the tribes in that the antennae are situated on prominent tubercles variously shaped and armed. As a rule the body is more or less naked, being covered only by a few scattered hairs. In the peculiar specialized Cervaphidina, however, large toothed processes extend outward from the body surface. Migration between a primary host and one or more secondary hosts often occurs. Apterous and alate viviparous forms, therefore, are common, but no definite relation exists between them. The forms feed mostly upon the leaves of trees and herbs but they may also be found feeding upon the twigs and roots. They are not infrequently attended by ants. The oviparous females are nearly always apterous, but the males, on the other hand, usually are winged. Apterous males, however, are common and intermediate forms between alate males and apterous males sometimes occur. Inter- 40 BULLETIN 826, U. S. DEPARTMENT OF AGEICULTURE. mediates between apterous and alatc viviparous forms are of quite common occurrence and indicate possible origin of the apterous forms. It is no doubt true, however, that in members of this tribe the equilibrium is disturbed more easily than in some of the others, and that external influences have a more sudden and noticeable effect. We have divided the tribe into four subtribes which may be separated as follows: Key to the Subtribes of the Aphidini. 1. Body covered with long projections Cervaphidina. Body naked with the exception of a few hairs 2. 2. Head without prominent antennal tubercles Aphidina. Head with prominent antennal tubercles 3. 3. Winga with the radial sector normal Macrosiphina. Wings with the radial sector more or less united with the upper branch of the media or hind wings reduced Pentalonina. Subtribe APHIDINA. The insects in the Aphidina show a great variation in regard- to the cornicles and cauda. Some have very markedly developed cornicles, others have extremely small ones, while one genus lacks them altogether. The cauda varies from very large in genera like Hyalopterus to scarcely any visible cauda in some of the other genera. Certain of the genera appear more similar than others, for example, Hyalopterus, Pergandeidia, and Brachycolus all have small cornicles and somewhat large caudas. Certain other genera, while appearing quite different in some ways, are evidently related. Cavariella, Hyadaphis, Aspidaphis, and Vesiculaphis all have char- acters which are very suggestive, although there are differences between thsm. So also there is a group suggesting Aphis. The various genera may be separated as follows: Key to the Genera of the Aphidina. 1. Cornicles absent Asiphonaphis. Cornicles present 2. 2. Cornicles swollen, not subcylindrical or tapering 3. Cornicles subcylindrical or tapering but eometimes extremely short and ringlike 12. 3. Fore wings with the media once branched; apterous form with very much swollen cornicles Vesiculaphis. Fore wings with the media twice branched 4. 4. Hind wings with the cubitus lacking Carolinaia. Hind wings with both media and cubitus present 5. 5. Abdomen with a dorsal projection or tubercle above cauda 6. Abdomen without this structure 7. GENERIC CLASSIFICATION OF APHIDIDAE. 41 6. Tubercle very large, entirely covering posterior part of body; cornicles email, opening at the side Aspidaphis. Tubercle of moderate size, as large as cauda in the apterous form; cornicles normal, opening at the end Cavariella. 7. Cornicles long, very abruptly and distinctly swollen Liosomaphis. Cornicles of varying lengths, gradually swollen 8. 8. Cauda short and abruptly conical; cornicles about the same length as Cauda and swollen in the middle Brevicoryne. Cauda not strikingly short or aljruptly conical 9. 9. Cornicles as short as the width of cauda at base or shorter 10. Cornicles as long as or longer than the cauda 11. 10. Cornicles minute, tubercle-like Brachycolus. Cornicles considerably longer than their diameter; cauda, particularly in the apterous form, large and long Hyalopterus. 11. Cauda in the apterous form long and l:)road, as long as the cornicles-.Hyadaphis. Cauda shorter than the cornicles, not correspondingly long and broad Rhopalosiphum. 12. Tarsi atrophied *. 13. Tarsi normal 14. 13. Antennae of six segments; cornicles rather short Atarsos. Antennae of five segments; cornicles very long and slender Mastopoda. 14 . Cauda apparently absent, or a mere rounded platelike structi^e 21. Cauda normal in appearance but often very short 15. 15. Fore wings with media once branched 22. Fore wings with media twice branched 16. 16. Antennae of five segments 24. Antennae of six segments 17. 17. Cornicles small, as short as width of cauda at base, which is mthout con- striction 18. Cornicles usually as long as or longer than cauda, Aphis-like 19. IS. Cauda very long and large in the apterous form Pergandeidia. Cauda extremely short and subcorneal Microsiphum. 19. Cauda short and abruptly conical Anuraphis. Cauda elongate and constricted near base 20. 20. Hind wings with media and cubitus present Aphis. Hind wings with only one oblique vein Hysteroneura. 21. Cornicles minute, not as long as wide Cryptosiphum. Cornicles moderate in length Acaudus. 22. Antennae of six segments Toxoptera. Antennfe of five segments 23. 23. Apterous form with a prominent median projection on the vertex and with four-segmented antennae; cornicles minute Sanbornia. Apterous form normal Yamataphis. 24. Cornicles elongate. Aphis-like .« Cerosipha. Cornicles extremely short, body of apterous form, much arched . Siphonatrophia, Genus ACAUDUS V. d. Gool. riateVI.A, B. 191.3. Acaudus Van dcr Goot, Tijd. voor Ent., v. 56, p. 97. 1917. Macchiatiella Del Guercio, Redia, v. 12, p. 210. 1917. Hannabura Matsumura, Joum. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 377. Van der Goot erected his genus Acaudus with lycTinidis L. as type. This species is quite similar to an Aphis without a cauda or with a very short, somewhat rounded cauda. The cornicles are not long. 42 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. D<.'1 Guercio's genus had his trifolii placed as t}']>e. The corni- cles of this species appear to be a little longer than those of the type of Van der Goot's genus, but in other respects the insects seem to be quite similar. We feel that they belong to the same genus. We are keeping this genus distinct from Cryptosiphum, not only because of the minute cornicles in that genus but also on account of the peculiar head structure which is there seen. It is not typically Aphis- Hke. Characters. — Head without prominent antennal tubercles. Antennae of six seg- ments. Fore wings with the media twice branched; hind wings with both media and cubitus present. Cornicles cylindrical, of moderate length: cauda reduced to a broad, short, rounded structure. Type (fixed by Van der Goot, 1913), Aphis lychnidis L. Genus ANURAPfflS Del Guercio. riate VI, C-F, 1907. Anuraphis Del Guercio, Redia, v. 4, p. 190. 1913. Brachycaudus Van der Goot, Tijd. voor Ent., v. 56, p. 97. 1913. Dentatus Van der Goot, Tijd. voor Ent., v. 56, p. 98. 1913. Scmiaphis Van der Goot, Tijd. voor Ent., v. 56, p. 103. 1917. Yezabura Matsumura, Journ. Coll. Agr. Toholni Univ., v. 7, pt. 6, p. 392. 1918. Sappaphis Matsumura, Trans. Sapporo Nat. Hist. Soc, v. 7, pt. 1, p. 18. In 1907 Del Guercio erected the genus Anuraphis in which were included 'pyri Koch, lappae Koch, iridis Del Guercio, ranuncuWKsdt., myosotidis Koch, centaureaeKoch., prunicola Kalt., tragopogonis Kalt., iani Peer, farfarae Koch, and persicae Boyer. These species, with the exception of lappae and tragopogonis which seem to belong to Aphis, have a broadly and somewhat abruptly conical cauda quite unlike that of the genus Aphis. In 1013, Van der Goot erected his genus Brachycaudus with one of these species, myoso- tidis, as type wdthout referring any of the species to Anuraphis at all. It is evident that Brachycaudus is a synonym of Anuraphis. At the same time he erected the genus Semiaphis with carotae Koch as type. A study of this species, as seen in Plate VI, makes it evident that this is a species of the same general type. Semiaphis thus becomes a synonym of Anuraphis. Based on the minute tubercles, particularly on the head and caudal portion of the abdomen. Van der Goot also erected the genus Dentatus with sorhi Kalt. as type. Apart from these tubercles sorhi is in the character of the cauda, etc., a ty])ical Anuraphis. In our American . rosy apnis, a species very similar to sorli, these tubercles are absent in many individuals, and in the fall migrants the caudal ones are nearly always absent. If this character were retained, therefore, the species would belong in one genus as far as the spring migrant is concerned, and in another genus when the fall migrant is considered. Dentatus, therefore, becomes also a synonym of Anuraphis. GENERIC CLASSIFICATION OF APHIDIDAE. 43 Characters. — Head without prominent antenual tubercles. Antennae of six seg- ments and armed with subcircular sensoria. Fore Avings mth the media twice branched, hind wings with both media and cubitus present. Cornicles cylindrical, often short, though sometimes moderately long; cauda short, broad, and abruptly conical, never elongate and constricted as in Aphis. Males usually winged. Ovi- parous forms apterous. Type (fixed by Del Guercio, 1907), Aphis pyri Koch. Genus APHIS L. Plate VI, G-I. 1758. Aphis Linnaeus, Systema Naturae, lOth ed., p. 451. 1817. Lozerates Rafmesque, Am. Mo. Mag. & Crit. Review, v. 1, p. 361. 1907. Uraphis Del Guercio, Redia, v. 4, p. 192. 1907. Microsiphon Del Guercio, Redia, v. 4, p. 192. 1913. MyzapMs Van der Goot, Tijd. voor Ent., v. 56, p. 96. 1913. StenapMs Del Guercio, Redia, v. 9, p. 185. 1916. Longiunguis Van der Goot, Zur Kenntniss der Blattliiuse Java's, p. 112. 1916. Melanaphis Van der Goot, Zur Kenntniss der Blattliiuse Java's, p. 61. 1917. Abvra Matsmnura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 407. 1917. Arimakia Matsuraura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 405. A number of species were included in the original genus by Lin- naeus. Of these Lamarck set ApMs ulmi L. as type in 1801, and in 1802 Latreille set ApMs sarribuci L. as type. ApMs ulmi L. is, according to Passerini, Eriosoma lanuginosa Hartig and the spe- cies now placed in Tetraneura. Samhuci is retained here as type and a request will be submitted that this species be fixed definitely by the International Commission. It would greatly' disarrange the economic literature to change the meaning of this common name. In 1907 Del Guercio erected two genera, Uraphis and Microsiphon based on the relative length of the cornicles and cauda. When con- sidering certain individual species this would appear as a very fair character for use. But when large series of species are studied it will be found that in the species having the cauda of the typical Aphis shape there are all gradations of cornicles from the very short to the very long. This will be seen also in the forms having the abruptly conical cauda. Some species have very short cornicles and some quite long ones. Moreover, in the same species the cornicles in the different forms will bear a different relation to the length of the cauda. Species, therefore, having cornicles and cauda of essentially the same character should not be used as types of different genera depending on the length of the cornicles. If this were the case, certain species which are close to the border line of separation would on some individuals fall in one genus and on other individuals fall in a different genus. Under this rule, saccliari Zehnt, would, it is believed, belong to the genus Aphis, and Longiunguis, therefore, be a synonym of Aphis. Likewise the genus Melanaphis would be a synonym of Aphis. This genus was erected with lanthusae Kirk, as type. The cornicles are short but the cauda is aphis-like. Where the generic characters are considered, the genus appears a synonym. 44 BULLETIN 826^ V. S. DEPARTMENT OF AGRICULTURE. The genus Myzoaphis was erected by Van der Goot with Aphis rosarum Kalt. and Aphis ahietina Walker. These two species show practically no antennal tubercles and are very little different from a typical Aphis, excepting in the clothing of the antennge and body. The Cauda as figured by Van der Goot is quite conical, but the writer's specimens are somewhat different from this and specimens of ahietinus do not show a cauda exactly like his figure. In fact, they appear more like an iVphis. This genus, therefore, should not be separated or it will necessitate the separation of very many other forms under new names. Characters. — Head without prominent antennal tubercles. Antennae of six seg- ments and armed with subcircular sensoria. Fore "wangs with the media twice branched ; hind winga with both media and cubitus present. Cornicles cylindrical or slightly tapering. Cauda usually not as long as the cornicles, subcorneal, rather elongate, constricted about the middle. Anal plate rounded. Males usually winged, oviparous females apterous. Type (by suspension of rules), Aphis samhuci L. Genus ASPIDAPHIS Gillette. Plate VI, L-0. 1917. Aspidaphis Gillette, Can. Ent., v. 49, p. 196. The genus Aspidaphis appears to be related both to Cavariella and to Vesiculaphis. It has the short, blocky form of Vesiculaphis and also the peculiar integument. On the other hand, certain species with a similar integument are met with in Cavariella. The develop- ment of the dorsal abdominal tubercle is here very pronounced and the cornicles have taken on a peculiar shape. Characters. — Head without prominent antennal tubercles; antennse short, of five segments, armed with subcircular sensoria. Wing venation normal. Cornicles very small, somewhat swollen near the distal extremity and with the opening in the side of the cornicle, not at the tip. Abdomen with a dorsal caudal tubercle developed into^ a large conical process extending lieyond and fully covei'ing the cauda in the apterous form. Body elongate. Type (monotypical), Asjndaphis polygonii Gill. Genus ASIPHONAPmS Wilson & Davis. 1919. Asiphonaphis Wilson A Davis, Ent. News, v. 30, p. 39. Characters. — Head without prominent antennal tubercles. Anteimse of six segments armed with subcircular sensoria. Fore wings with the media twdce branched; hind wings with both media and cubitus present. Cornicles absent entirely. Abdomen with large lateral tubercles. Catida- somewhat conical or Aphis-like. Anal plate- rounded. Type (monotypical), Astphonaphh prniil Wilson <& Davis.. Genus ATARSOS Gillette. I'Mc VI, r-s. 1911. Atarsos Gillette, Ent. News, v. 22, p. 440. Characters. — Head without prominent antennal tubercles. Antennie of six segments, armed with subcircular, somewhat tuberculate sensoria; fore wings with the media twice branched, hind winga with both media and cubitus present. Cornicles rather GENERIC CLASSIFICATION OF APHIDIDAE. 45 short, subcylindrical; cauda somewhat conical, not quite as long as the cornicles; anal plate rounded. Tarsi atrophied in all the forms. Type (monotypical), Atarsos grindeliae Gill- Genus BRACHYCOLUS Buckton. Plate VI, T, U. 1879. Brachycolus Buckton, Mon. British Aphides, v. 2, p. 146. 1913. Brachysiphum Van der Goot, Tijd. voor Ent., v. 56, p. 105. The genus Brachycolus was erected with stcllariae Hardy as type and is easily recognized from the structure of the cornicles. Van der Goot's genus is in all essential respects the same, the type of that genus being thalictri Koch. Characters. — Head without prominent antennal tubercles. Antennse of six segments and armed with subcircular sensoria. Wing venation normal. Cornicles very small, especially in the apterous form. Cauda medium in size and conical. Body elongate; legs and antennae usually short. Type (monotypical), Aphis stellariae Hardy. Genus BREVICORYNE V. d. Goot. I'late VI, J, K. 1915. Brevicoryne Van der Goot, Beitrage z. Kennt. d. HoU. Blattlause, p. 245. 1916. Oedisiphum Van der Goot, Zur Kenntniss der Blattlause Java's, p. 122. 1918. Brevicoryne Das, Mem. Ind. Mus., v. 6, p. 179. The genus Brevicoryne was erected with Ai?M.s hro.ssicae L. as type, a species in which the cornicles are very short and somewhat swollen in the middle, and the cauda conical. In the species Oedisiphum com'positarum V. d. Goot quite similar characters are found. The cornicles appear to be somewhat more slender but in the main the species appear alike. Oedisiphum, therefore, becomes a synonym. Characters. — Head without prominent antennal tubercles; antennse of six segments and armed with subcircular sensoria. Wing venation normal. Cornicles short, not much longer than the cauda and swollen in the middle. Cauda short and broadly conical. Forms not especially elongate. Type (monotyj^ical), Aphis hrassicae L. Genus CAROLINAIA Wilson. Plate VI, V, W. 1911. Carolinaia Wilson, Can. Ent., v. 43, p. 61. The genus Carolinaia Wilson is related very closely to Rhopalosi- phum. In fact some of the species placed here are distinguished from members of that genus only by the fact that the cubitus is lacking in the hind wing. It is true that the type species was described as having five-segmented antennse in the apterous form. But other species, like this in all other respects, have normal six- segmented antennae. Certain species of Rhopalosiphum will show strains in which nearly all of the individuals will have five-segmented antennse and yet the normal antenna for such species is a six- segmented one. In this closely related genus, therefore, it is not surprising if a similar condition is met with. 46 BULLETI^Sr 826^ U. S. DEPARTMENT OF AGRICULTURE. Characters. — Head without prominent antennal tubercles. Antennae of five or six segments armed with subcircular sensoria. Venation of the fore wings normal; hind wings with the cubitus absent; cornicles elongate, slightly swollen near the distal extremity, but Avithout a prominent neck near the proximal end. Cauda rather broadly conical. Type (monotypical), CaroUnaia caricis Wlsn. Genus CAVASIELLA Del Guercio. Plate VI, X-Z. 1911. Cavariclla Del Guercio, Redia, v. 7, p. 333. 1914. Corynosiphon Mardwilko, Faune de la Russie Insecta, Aphidodea, p. 73. 1917. Nipposiphum Matsumura, Joum. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 410. Two new genera were erected in recent years: One by Del Guercio, Cavariella, with pastinacae L. as type; the other by Mordwilko, called Corynosiphon, but with no species definitely given. In a footnote, however, Mordwilko refers capreae Fab. to his Corynosiphon. Both of these genera, therefore, were used for insects of the same general type. This may be said too of Nipposiphum. Characters. — Head without distinct antennal tubercles; antennae of six segments, armed with prominent sensoria. Wing venation normal. Cornicles somewhat swollen near the distal end; cauda rather elongate, somewhat conical; abdomen with a tubercle or horn above the cauda, this tubercle or horn most prominent in the apterous form. Males winged; oviparous females usually apterous. Type, Aphis pastinacae L. ' Genus CEROSIPHA Del Guercio. 1900. Cerosipha Del Guercio, Nuove Rel. Staz. Firenze, sen 1, no. 2, p. 116. 1918. Metaphis Matsumura, Trans. Sapporo Nat. Hist. Soc, v. 7, pt. 1, p. 1. The genus Cerosipha Del Guercio was erected for a species, pas- seriniana Del Guercio, somewhat similar to Aphis but with five seg- ments to the antennae. The writer has been unable to study the type species and has based his remarks on ruhifolii Thomas, an American species. Characters. — Head without prominent antennal tubercles; antennae of five segments, Aphis-like, armed with subcircular sensoria. Fore wings with the media twice branched, hind winga with both media and cubitus present. Cornicles cylindrical or somewhat tapering, cauda Aphis-like, somewhat tapering. Tj-pe (monotypical), Cerosipha j^nsseriniana Del G. Genus CRYPTOSIPHUM Buckton. Plate VI, PP, QQ. 1879. C'ryptosiphum Buckton, Mon. Br. Aphides, v. 2, p. 144. The genus Cryptosiphum Buckton can be distinguished from others having very short cornicles l)y the cauda which is here short and rounded, while in most other cases it is very long. Characters. — Head without distinct antennal tubercles. Antennae of six segments, rather short, and with subcu'cular sensoria. Fore wings with the media either once or twice branched. Hind wings with both media and cubitus present. Cornicles sub- 1 The writer has a record to the effect that in a published paper Del Guercio used pastinacae as type of Cavariella, but he has been unable to locate the publication and can not reach Doctor Del Guercio through the mails. GEFEEIC CLASSinCATIOISr OF APHIDIDAE. 47 cylindrical but extremely short, not as long as wide. Cauda very short and rounded, not Aphis-like; anal plate rounded. Type (monotypical), Cryptosiphum artem isiae Buckt Genus HYSTERONEURA Davis. 1919. Heteroneura Davis, Can. Ent., v. 51, p. 228. 1919. Hysteroneura Davis, Can. Ent., v. 51, p. 263. Characters. — Head without prominent antennal tubercles. Antennae of six seg- ments armed with subcii'cular sensoria. Fore wings with the media twice branched. Hind wings with the cubitus absent. Cornicles somewhat tapering or subcylindric, Cauda Aphis-like; anal plate rounded. Type (monotypical), Aphis setariae Thos. Genus HYADAPHIS Kirk. Plate VI, AA, BB. 1904. Hijadaphis Kirkaldy, The Entomologist, v. 37, p. 279. 1863. Siphocoryne Passerini, Aphididae Italic*, p. 8 (not Siphocoryne , 1860). As indicated under the discussion of Siphocoryne, Passerini set xylosiei as type of his genus in 1863. Nymphaeae h-ad, however, been set in 1860, so Kirkaldy gave Hyadaphis to Passerini 's 1863 conception, of which genus xylostei becomes the type. Xylosiei has no caudal horn and is quite similar in general appearance to a Khopalo- siphum. We may separate the two genera, however, on the cauda, which in xylostei, particularly in the apterous form, is very large, fully as long as the cornicles, and broad, quite unlike that of nym'ph.aeae- Several other species which have generally been considered in the same genus with xylostei possess a distinct caudal projection on th# abdomen. One of these has been made the type of Cavariella so that such species will be removed from our conception of Hyadaphis. Characters.— Wi.Q&di without prominent antennal tubercles; antennae of six segments which are rather abundantly armed with tuberculate sensoria. Wing venation normal. Cornicles somewhat swollen but not prominently so. Cauda, particularly in the apterous form, large, as long as the cornicles, and broad. Males usually winged; oviparous females apterous; summer forms usually feeding on the Umbelliferse. Type (fixed by Kirkaldy, 1904), Aphis xylostei Schrk. Genus HYALOPTERUS Koch. Plate VI, RR-WW. 1854. Hyalopterus Koch, Die Pflanz. Aphiden, p. 16. 1917. Hayhurstia Del Giiercio, Redia, v. 12, p. 208. Apliis iwiini Fab. was set as the type of this genus in 1860 by Passerini. In 1917 Del Guercio erected the genus Hayhurstia. There is little difference between the two as will be seen by examining the drawings given herewith. The cauda in Hayhurstia is slightly nairower than that in Hj'alopt^rus. It is the wi'iter's opinion that the two represent only one genus. Two other genera may be mentioned here. They are Brachycolus Buckt. and Pergandeidia Schout. Specimens of the type species of Brachycolus show that this genus is quite similar to Hyalopterus, but it differs in the cornicles. In Brachycolus the cornicles are very small, almost mere rounded swellings in the apterous form, whereas 48 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. in Hyalopterus the cornicles are in the apterous form of fair size and not distinctly swollen. Specirnens of the type of Pergandeidia received from Schouteden show that this genus is very close to Hyalopterus and is probably almost too close for a very distinct genus. The diagnosis given by Wilson (1910) for this genus does not agree with the type species as determined by the author of the genus. In the specimens examined b}^ the writer the cauda, as will be seen in the drawings, is very much longer than the cornicles, bearing about the same ratio as seen in Hyalopterus. Characters. — Head without distinct antennal tubercles, antennae of six segments armed Trith subcircular sensoria. Wing -s'enation normal. Cornicles Aery short, not much longer than the cauda is wide at its base, swollen beyond the middle, particularly in the alate form. Cauda long and broad, considerably longer than the cornicles. Form of the insects elongate, often more or less fiat. Type (fixed by Passerini, 18G0), Aphis pruni Fab. (=A. arundinis Fab.). Genus LIOSOMAPHIS Walker. Plate VI, NN, GO. 1868. Liosomaphis Walker, The Zoologist, p. 1119. The genus Liosomaphis Walker is related somewhat closely to lihopalosiphum. The two genera, however, can be separated on the structure of the cornicles. In Liosomaphis the cornicles have a very distinct neck near the proximal extremity, due to a constriction behind the prominent swelling. This is strikingly evident in the apterous form, as well as in the alate one. In Rhopalosiphum, on the other hand, there is no abjupt swelling, but only a gradual one which is not at all prominent, as in Liosomaphis. Characters. — Head without prominent antennal tubercles. Antennae of six segments armed with subcircular sensoria. Wing venation normal. Cornicles elongate, distinctly swollen in the middle, and with a constricted neck near the base. Structure in both the apterous and alate forms similar. Cauda not as long as the cornicles, somewhat narrowly conical. Males usually winged; oviparous females usually apterous. Type (monotypical). Aphis herheridis Kalt. Genus MASTOPODA Oestlund. 1886. Maslopoda Oestlund, Minn. Oeol. Surv. Kept. 14, p. 52. The genus Mastopoda Oestlund, like Gillette's Atarsos, is peculiar in that the tarsi are atrophied. Characters. — Head Avithout distinct antennal tubercles. Antennae of five segments. Fore wings with the media twice branched, hind wings with both media and cubitus present. Cornicles somewhat long and cylindrical. Cauda short, conical, Aphis- like. Legs with the tarsi absent and provided instead with a membranous disk which enables the insect to walk inverted on smooth surfaces. Type (monotypical), Mastopoda pteridis Oestlund, GENERIC CLASSIFICATION OF APHIDIDAE. 49 Genus MICROSIPHUM Cholodkovsky. 1008. Microsiphum Cholodkovsky, Zool. Anz., v. 32, p. 087. Characters. — Head without prominent antennal tubercles, although with apparent ones. Antennae of six segments, rather long and slender. Fore wings with the media twice branched ; hind mngs with both media and cubitus present. Cornicles very- short, not much longer than wide and often scarcely Aasible in the alate form. Cauda extremely short and subcorneal or slightly rounded; anal plate rounded. Type (monotypical), Microsiphum ptarmicae Choi. Genus PESGANDEIDIA Schoutedgn. rbte VI, KK-MM. 1903. Pergandcidia Schouteden, Zool. Anz., v. 20, p. 085. 1913. Longicaudus Van der Goot, Tijd. voor Ent., v. 50, p. 140. 1915. RMzoberlesia Del Guercio, Redia, v. 10, p. 240. 1918. Yezosiphum Matsumura, Trans. Sapporo Nat. Hist. Soc, v. 7, pt. 1, p. 7. 1918. Brachyunguis Das, Mem. Ind. Mus., v. 0, p. 227. The genus Pergandeidia was erected by Schouteden with his ononidis as type. The characterization of the genus in this paper is based on specimens of the species received from Schouteden, and though it does not agree with the characterization of the genus as sometimes given, it seems necessary to follow the type species in forming a conception of the genus. In 1913 Van der Goot erected the genus Longicaudus and placed trirliodus Walker as type. In studying this species differences sufficient for good generic distinc- tion have not been found, and, therefore, the conclusion may be drawn that Longicaudus is a synonym of Pergandeidia. Del Guercio's genus is placed here also, although a study of the type species, which is not available, possibly may show the cauda to be somewhat different. Characters. — Head with no prominent antennal tubercles; antennse of six segments. Fore wings with media twice branched, hind wings with both media and cubitus present. Cornicles very short, almost as wide as long. Cauda very long and broad. Type (monotypical), Pergandeidia ononidis Schout. Genus RHOPALOSIPHUM Koch. Plate VI, FF-JJ. 1854. Khopalosiphum Koch, Die Pflanz. Aphiden, p. 23. 1860. Siphocoryne Passerini, Gli Afidi, p. 28. 1882. Rhopalosiphon Scudder, Nomenclator Zoologicus. 1910. Coloradoa Wilson, Ann. Ent. Soc. Am., v. 3, p. 323. 1915. SipUonapMs Van der Goot, Beitrage zur Kennt. d. Holl. Blattliiuse, p. 238. 1918. Stephensonia Das, Mem. Ind. Mus., v. 6, p. 175. 1856. (Type fixation) Gerstaecker, Bericht for 1S54, p. 102. There has been much confusion in regard to this genus. Koch included a number of diverse species. In 1856 Gerstaecker defi- nitely set Aphis nymplieae L. as type. Overlooking this, Passerini in 1860 set persicae Sulz. as type. In 1863, however, he changed this name to dianthi Schrank. In 1860 Passerini erected the genus Sipho- coryne with A. nymplieae L. as type. Having the same type this must, therefore, become a synonym of Rhopalosiphum. In 1863 Passerini placed nymplieae in Rhopalosiphum, although he set dianthi 141613°— 20— Bull. 826 4 50 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. as type, and he used Siphocoryiie in a different sense with xylostei Schrank as type. Such a procedure is inadmissible, and Kirkaldy therefore renamed this genus Hyadaphis. The genus Colorado a Wilson was erected with rufomaculata Wilson as type. Although much smaller, this species is essentially like nympheae in structure, and we therefore consider the genus a syno- nym of Rhopalosiphum. The genus Siphonaphis Van der Goot was erected with nympheae h. as type. Having the same type, therefore, it must become a synonym of Rhopalosiphum. Lahorensis Das is quite similar. Characters. — Head without prominent antennal tubercles. Antennae of six seg- ments with the usual subcircular sensoria present. Wing venation normal. Cor- nicles moderately long and slender, slightly swollen near their distal extremities. Cauda rather elongate, not as long as the cornicles, and not broad. Abdomen without a caudal horn or projection above the cauda. Males usually winged; oviparous females usually apterous. Type (fixed by Gerstaecker, 1856), Aphis nympheae L. Genus SANBORNIA, n. gen. riate Vn, F-L. The genus Sanbornia is erected for a peculiar form living on juniper at College Station, Tex., and forwarded to the Bureau by Charles Sanborn. This species was determined as undescribed by Mr. Pergande who had planned to j^ublish on it. He had given it the name juniperi. The type is in the U. S. National Museum (Cat. No. 23064). Characters. — Head without prominent antennal tubercles. Atitennse of five seg- ments, armed with circular sensoria. Fore wings with the media once branched, hind wings with only the media present, cornicles minute; cauda elongate. Apterous fonn with four-segmented antennse and with a prominent mushroom-like projection on the vertex. Type, Sanbornia juniperi Perg. Sanbornia jiiniperiVergiLude, n. sp. (Description by Mr. T. Pergande.) Apterous form. — The head is most remarkable in front, having a large, squarish, bilobed projection about the middle and each side of it; close to the insertion of the antennae is a prominent, short, and conical protuberance. There is also at the inner side of the first antennal joint a long and slightly conical protuberance. The antennae are but four-jointed, the spur shorter than the basal section of the joint. Nectaries are not visible (?); the tail is rather long and uniformly elongate conical; the tarsi are very short, the first joint appears to be minute, and in alcoholic specimens seems to be withdrawn into the tibise; the last abdominal segment is semicircular. GENERIC CLASSIFICATIOI^ OF APHIDIDAE. 51 Genus SIPHONATROPHIA Swain. Plate VII, A-E. 1918. Siphomtropliia Swain, Ent. News, v. 29, p. 363. Characters. — Head without prominent antennal tubercles. Antennae of five seg- ments and armed with circular sensoria. Fore wings with the media twice branched; hind wings with both media and cubitus present. Cauda elongate and conical. Cor- nicles extremely short. Body of apterous form considerably arched. Type (monotypical), Cerosipha cupressi Swain. Genus YAMATAPHIS Matsumura. 1917. Yamataphis Matsumura, Jour. Coll. Agr. Tohoku Univ., v. 7, pt. 6, p. 412. Characters. — Head without prominent antennal tubercles. Antennae of five seg- ments armed with small circular sensoria. Fore wings with the media once branched; hind wings with both media and cubitus present. Cornicles subcylindrical. Cauda Somewhat conical. Type (fixed by Matsumura, 1917), Yamataphis orijzae Mats. Genus TOXOPTERA Koch. Plate VI, DD, EE. 1857. Toxoptera Koch, Die Pflanz. Aphiden, p. 253. 1S91. Ceylonia Buckton, Ind. Mus. Notes, v. 2, p. 3.5. Kooh's genus was erected for his aurantiae, a weU-known species on citrus, etc. Buckton erected his genus for a species he described as tlieaecola. Specimens of aurantiae from various regions and specimens of tlieaecola from Zehntner substantiate the placing of tlieaecola as a synonym of aurantiae. Ceylonia will then become a synonym of Toxoptera. Even if the two species were held to be distinct, this would necessarily be the case. Characters. — Head without prominent antennal tubercles. Antennae of six seg- ments, armed with subcircular sensoria. Fore wings with the media once branched, hind wings with both media and cubitus present. Cornicles moderate in length, sub- cylindric, tapering. Cauda of moderate length, somewhat constricted near the base. Type (monotypical), Toxoptera aurantiae Koch {aurantiae Boyer). Genus VESICULAPHIS Del Guercio. Plate VII, M-Q. 1911. Vcsiculaphis Del Guercio, Rcdia, v. 7, p. 4(54. Characters. — Apterous form elongate ; antennae short, composed of five segments and situated on the under side of the head . Top of head forming a ledge which extends out over the antennse forming an angle in fi'ont of the eye; eyes protruding. Cornicles large, very much swollen and curved, opening minute and flanged; cauda somewhat conical, rounded at the tip. Anal plate rounded; posterior part of abdomen extending out over the anal plate and somewhat over the cauda. Alate form with six-segmented antennte armed with subcircular tuberculate sensoria. Fore wings with the media once l:)ranched, hind wings with both media and cubitus present. Cornicles somewhat slender, swollen, and slightly constricted at the tip. Cauda conical, not as long as the cornicles, anal plate rounded. Type (monotypical), Toxoptera caricis Fullaway. This genus is one of those peculiar genera, Uke Aspidaphis Gillette, which seem to be related to Rhopalosiphum. 52 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. Subtribe CERVAPHIBINA. The subtribe Cervaphidina, is a most interesting ami peculiar one. Tlie specialization is remarkable in that long processes are developed on the body and considerable reduction has taken place in the wings and antenna? while no Aphis-like cauda is found. Only two genera are known, which may be separated as follows: Key to the Genera of the ('era^at'hidina. Cornicles swollen; body spine-like projections not armed with teeth. . . Anomalaphis. Cornicles not swollen; body spine-like projections armed with teeth Cervaphis. Genus ANOMALAPHIS, n. gen. I'latc VIII, D-F. Characters. — Body armed with elongate tubercle-like projections, particularly on the caudal portion; antennae five-segmented in bcth apterous and alate forms, armed with subcircular sensoria. Fore wings with the media once branched; hind wings considerably reduced, with the cubittis absent. Cornicles distinctly swollen; cauda and anal plate reduced. Type Anomalaphis comperei Pergande. Anomalaphis comperci Pergande, n. sp. Among the many descriptive notes left by Mr. Theo. Pergande are some recording a peculiar species from Australia. This proves to represent an undescribed genus in the Cervaphidina. Pergande rec- ognized the species as typical of a new genus to which he gave the manuscript name here used. He left no description of the genus and his notes on the species are given here exactly as he left them. The type is in the United States National Museum collection of Aphididae (Cat. No. 23065). Feb. 18, 1907, Rec. from Compere, a lot of Aphides, found in 1901, on Acacia and Eucalyptus, along the beach at Albany, West Australia, which represents a new genus among Rhopalosiphius, and is a most remarkable Aphid in various respects. The antennfo, in the apterous and migratory female, are but 5-jted . while the spur is rather short and resembles that of Chaitophorus. The front wings are ample and reach con- siderably beyond the end of the body, with the third discoidal having but one fork, as in Schizoneura. The hind wings are very short and narrow and reach out to the apex of the 1st vein of the anterior wings, there is also but 1 discoidal, straight, and near the apex of the wing. The nectaries are short, clavate, and similar to those of Siphocoryne, the tail appears to be wanting. The abdomen of the migrant appears to have been of a dusky yellowish green, with transverse rows of small, black spots or tubercles, and blackish sutiu-es between the segments. The eyes are brown; antennae black, rather short, reaching barely to the abdomen and l)ut 5-jointed; the two basal joints as usually; the 3rd joint is longest, about as long as the remaining joints together, including the spur, Avith some projecting sensoria and a few short bans; joints 4 and 5 are subequal in length, exclusive of the spur, and clavate, the spur is about ^ the length of the basal section of the joint, rather stout and blunt; the front of the head resembles that of Aphis. The sides of the abdominal segments are somewhat angulated, each angle pro\'ided with a very short, capitate, stout l)ristle, while at the posterior edge of the two segments, following the nectaries, there is a pair of long, GENERIC CLASSIFICATION OF APHIDIDAE. 53 diverging, fleshy spines, with a sharp, quite long and slender spine at the tip; the posterior pair longest; all of them black; the end of the body is fringed "with fine and quite long hau's. Legs as in other Ai)hides. The apterous females are dark brownish or grayish green above, with a somewhat fusi- form median, yellowish strip, broadest near the head , taj^ering posteriorly to a point and terminating in front of nectaries; the sutures of the segments, the sides and under side of the body are also of a yellowish color, on account of which, there is each side a subdorsal row of transverse, dark spots. The head and about basal half of the antennae, dark, dirty yellowish, the eyes dark brown. There are about 4 short and curved capitate hairs on the front of the head and prominent fleshy tubercles each side of the body, each bearing at its apex a short, capitate spine or hair, all of them growing longer toward the nectaries, while beyond the nectaries there are two pairs of long and slender fleshy tubercles, tipped with a spine, ?s in the migrant. A tail could not be seen. In the younger forms and pupae, the tubercles are as in the apterous female. In the pupae the head, prothorax, abdomen and nectaries are of a dkty yellowish color, with transverse rows of small, black or dusky spots on the abdomen. The wing pads are black. Genus CERVAPHIS Van der Goot. Plate VIII, G. 1916. Ccrvaphis Van der Goot, Ziir Keiuitniss der Blattlilusc Java's, p. 14S. Characters. — Body armed with a series of long toothed projections; antennae of apterous form five-segmented, of the alate form six-segmented with somewhat oval sensoria. Fore wings with the media once branched; hind wings greatly reduced in size and lacking both the media and cubitus; cornicles elongate, subcylindric. Cauda and anal plate reduced. Oviparous females often winged. Type (fixed by Van der Goot, 1916), Ccrvaphis schoutcdeniae V. d. Goot. Subtribe MACROSIPHINA. The genera of the subtribe Macrosiphina may be separated at once from those of the Aphidina in that they have developed large antennal tubercles. These may assume various shapes by which the genera often may be separated. Considerable variation is met with also in the cornicles and cauda, although as a rule the cornicles are very well developed. They may be either cylindrical or swollen. In one genus, Hyalopteroides Theo., the cornicles are very short, suggesting some of the genera of the Aphidina. The cauda is as a rule rather long. The wings are in nearly every case normal in venation. The different genera may be separated by the following key: Key to the Genera of the Macrosiphina.' 1 . Cornicles swollen 2. Cornicles cylindrical or tapering, scarcely swollen G. 1 Since this paper was set up Takahashi (Insect World, v. 23, p. 439) has erected the following genus, which will fall in this subtribe. Genus AKKAIA Takahashi. CAarac/crs.— Cornicles swollen; frontal tubercles and first antennal segment with prominent projec- tions. Antennee of five segments. Cauda somewhat knobbed, anal plate large and projecting. Type (monotypical), ^fcfcaia poZj/^fne, Takahashi. 54 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. 2. Cornicles cylindrical at base and extremity; abruptly swollen in middle. . Rhopalosiphoni nus. Cornicles with the swelling gradual 3. 3. Head with a large central process on vertex Francoa. Head without this - - 4. 4. Antennal tubercles large and diverging 5. Antennal tubercles converging; head and basal antennal segments with very prominent capitate hairs Capitophorus. 5. Cornicles much longer than cauda which is somewhat tapering. . Amphorophora. Cornicles about the length of cauda M'^hich is usuall}^ constricted near its base Megoura. 6. Cornicles very small, much smaller than the long, broad cauda. .Hyalopteroides. Cornicles as long as or longer than the cauda 7. 7. Head with prominent, elongate projections to the antennal tubercles, par- ticularly eA^dent in the apterous form Phorodon. Head without these 8. 8. Tubercles strongly converging ; Myzus. Tubercles distinctly diverging 9. 9. Cornicles thick, about as long as the cauda which is large and somewhat constricted near base Macrosiphonella. Cornicles very long, rather slender, subcylindric, somewhat tapering 10. 10. Cauda elongate, constricted near base Macros! phum. Cauda moderate or elongate, not constricted near base 11 . 11. First antennal segment and abdominal segments with long fingerlike tuber- cles in the apterous form Acanthaphis. Without these Illinoia. Genus ACANTHAPHIS Matsumura. Plate VII, R-U. 1918. Acanthaphis Matsumura, Trans. Sapporo Nat. Hist. Soc, v. 7, pt. 1, p. 15. The genus Acanthaphis Mats, is somewhat related to Phorodon. Characters. — Head with prominent diverging antennal tubercles. Antennae of six segments, first segment with a long fingerlike projection in the apterous form. Corni- cles long, slender, and cylindrical. Cauda elongate, conical. Dorsum of abdomen with long fingerlike tubercles. Body with capitate hairs. Type (fixed by Matsumura, 1918), Acanthajikh rubi Mats. Genus AMPHOROPHORA Buckton. Plate IV, A, B. 1876. Amphorophora Buckton, British Aphides, v. 1, p. IS". 1886. Macrosiphum Ocstlund, Minn. Geol. Survey Rept. 14, p. 27. 1900. Macrosiphum Del Guercio, Nuove Rel. Staz. Fircnzc, ser. 1, no. 2, p. 159. 1901. Nectarosiphon Schoutcden, Ann. Ent. Soc. Belg., v. -15, p. 112. 1913. Eunectarosiphon Del Guercio, Rcdia, v. 9, p. 188. 1913. Rhopalosiphum Van dcr Gootj Tijd. voor Ent., v. 56, p. 146. Buckton erected his genus Amphorophora for his ampuUafa which he had secured in the apterous form only. Oestlund gave the generic name Macrosiphum to a species he described as ruhicola. In his sec- ond paper Oestlund describes a species under the name ampullata Buckton and says:'^ "The length of the antennae, together with the distinct frontal tubercles, may justify our exception of Ampho- rophora as a good genus." In speaking of his Macrosiphum • Bui. 4, Geol. and Nat. Ilist. Surv. Minn., p. 77, 1887. GENERIC CLASSIFICATION OF APHIDIDAE. 55 Oestlund says it may be too close to Rhopalosiphum and compares it with nym,pliaeae. But had he compared it with amjpullata which he also included in Rhopalosiphum he would probably have found it quite similar, and it is the writer's belief that Oestlund's species is in reality Amphorophora. Finding that Macrosiphum was preoccupied Schouteden gave the new name Nectarosiphon to rubicola and Nectarosiphon, therefore, will become a synonym. In 1913, Del Guercio erected the genus Eun^ctarosiphon with rubi Kalt. as type. There appears not to be sufficient difference here, however. The insects are in their main points the same and Eunectarosiphon is f)laced, consequently, as a synonym. Van der Goot in the same year used ampuUata as the type of Rhopalosiphum, In 1900 Del Guercio used the name Macrosiphum for a genus including three species: Convolvuli Kalt., viciae Kalt., and vubi'KiiM. He has since used both viciae and rubi as the types of other genera. This leaves only convolvuli in his Macrosiphum, and this species to all appearances is an Amphorophora. His Macrosiphum, therefore, is listed here under Amphorophora. Characters. — Head with promineut and slightly diverging antennal tubercles. Antennae of six segments, armed with circular sensoria. Wing venation normal. Cornicles long and somewhat swollen in the middle. Cauda elongate but much snorter than the cornicles. Type (monotypical), Am2:)horop1iora ampullata'Buii'kton. Genus CAPITOPHORUS Van der Goot. Plate VIII, A-C. 1913. Capitophorus'Vo.n dor Goot, Tijd. voor Ent., v. 5(3, p. 84. In the genus Capitophorus Van der Goot certain species suggest Rhopalosiphum in general characters, whereas others more nearly approach Myzus in their main characters. The capitate spines appear to be the best means of determining the genus. Characters. — Head with antenna! tubercles which are not markedly prominent, these each mth one or more prominent knobbed spines. Vertex with a central pro- jection on which similar spines are located. Antennae of six segments, armed with subcircular sensoria, the first segment with a projecting process on which one cr more knobbed spines are located. Wing venation normal; cornicles long and slender, slightly constricted in the middle and somewhat enlarged toward the distal extremity. Cauda rather short and conical. Type (fixed by Van der Goot, 1913), Aj)his carduinus Walker. Genus FRANCOA Del Guercio.i 1917. Francoa Del Guercio, Redia, v. 12, p. 201. The genus Francoa Del Guercio appears very close in some respects to Capitophorus but on account of the peculiar frontal tubercle and the structure of the first antennal segment it is held to be distinct. 1 It seems doubtful if this genus is distinct from Capitophorus. We have been unable to make a careful study of the tjq^e. 56 BULLETi:^' 826_, V. S. DEPARTMENT OF AGEICULTUEE. Characters. — Head with antenna! tubercles present. Vertex with a prominent rec- tangular process, both the antennal tubercles and the frontal process armed with knobbed spines. Antennas of six segments, the first segment lacking the process and capitate haii-s of Capita pliorus. Cornicles rather slender and swollen near the distal extremity. Cauda elongate, somewhat conical. Type (monotypical), Francoa elegans Del Guercio. Genus HYALOPTEROIDES Theobald. 1916. Hyalopieroides Theobald, The Entomologist, v. 49, p. 51. The genus Hyalopteroides was erected by Theobald for his species imUida found in the nest of Lasius niger, Porlock Weir, Somerset. It bears a striking resemblance to Pergandeidia but there are no prominent antennal tubercles in that genus. However, there are slight swellings suggestive of those figured by Theobald. The writer has never seen specimens of pallida and therefore is unable to give a personal opinion. Theobald says, "Head with marked frontal tubercles." This would place the genus as not closely related and pending a study of specimens it may be left thus. Characters. — Head with prominent antennal tubercles. Antennae of six segments and armed with subcircular sensoria. Comielea subcylindric, short, much shorter than cauda. Cauda long and conical. Type (monotypical), ITi/alopteroides pallida Theo. Genus ILLINOIA Wilson. Plate VIII, H-J. 1910. niinoia Wilson, Ann. Ent. See. Am., v. 3, p. 318. 1914. Mctopeurum Mordwilko, Faune de la Russie, Hemiptera, , v. 1, p. 56, 67. 1914. Acyrthosiphon Mordwilko, Faime de la Russie, Hemiptera, v. 1, p. 55, 62. This genus is closely related to Macrosiphum Pass, from which it may be distinguished by the nature of the cauda. Characters. — Head with prominent diverging frontal tubercles. Antennae of six segments armed Avith subcircular sensoria. Fore wings with the media twice branched, hind wings Avith both media and cubitus present. Cornicles cylindrical, sometimes slightly larger toward the middle which appearance is accentuated by a constriction often present near the distal extremity. Cauda conical, not as in JIacrosiphum with a constriction near its base. IMales usually mnged, o\iparous females ajjterous. Type (fixed by Wilson, 1910), Siphonophora liriodendri Mon. Genus MACROSIPHONIELLA Del Guercio. Plate VIII, R-T. 1911. ,1/oero.s/ptoraieKa Del Guercio, Rcdia, V. 7, p. 331. 1913. Macrosiphum Van der Goot, Tijd. voor Ent., v. 56, p. 145. 1914. Dielcysmura Mordwilko, Faune Russ. Aphidodea, p. 56. The genus Macrosiphoniella was erected by Del Guercio with atrum Ferr. as type. In 1913 Van der Goot used Macrosiphum of Passerini, indicating millefolii Fab. as type. This species can not be made the type of Macrosiphum Pass., and since it is essentially like atrum, V. d. Goot's Mascrosiphum must become a synonym of Macrosipho- niella. In 1914, Mordwilko used the generic name Dielcysmura for ■millefolii and figured the species. This name then proves also to be a synonym. GElsTERIC CLASSIFICATIOISr OF APHIDIDAE. 57 Characters. — Head with prominent diverging antennal tubercles. Antennae of six segments armed with suboircular prominent sensoria. Fore wings with the media twice branched, hind wings mth both media and cubitus present. Cornicles short, about the length of the cauda and rather thick, usually with conspicuous polygonal markings; cauda large, slightly constricted near the base. Type (fixed by Del Guercio, 1911), Aphis atrum Ferr. Ge"us MACROSIPHUM Passerini. Plate vni, U-W. 1860. Macros iplium Passerini, Gli Afidi, p. 27. 1855. Siphonophora Koch, Die Pflanzonlaiise Aphiden, p. 150. 1887. NectarOt-iiura Oestlimd, Minn. Geol. Survey Bui. 4, p. 78. 191.3. Macrosiphon Del Guercio, Redia, v. 9, p. 188. In 1855 Koch used the generic name Siphonophora for this genus for which Passerini substituted Macrosiphum. Not aware of this Oestlund, seeing that Siphonophora was preoccupied, substituted Nectarophora and strangely enough used Macrosiphum for another genus in a different sense. Del Guercio uses Macrosiphon in this sense. Characters. — Head with prominent diverging frontal tubercles. Antennse of six segments, armed with subcircular sensoria. Fore wings with the media twice branched ; hind wings with both media and cubitus present. Cornicles long, subcylindrical, mostly somewhat tapering; cauda long, somewhat constricted about the middle. Sexes with the male usually alate and the oviparous form apterous. ' Type (fixed by Passerini, 1860), Aphis rosae Linn. Genus MEGOURA Buckton. 1876. Megoura Buckton, British Aphides, v. 1, p. 188. 1913. Drepaniella Del Guercio, Rcdia, v. 9, p. 188. The genus Megoura Buckton is similar to jVmphorophora in that prominent antennal tubercles are present. It differs, however, in that the cornicles are short, about equal in length to the cauda. Buckton erected the genus for his viciae which Schouteden con- sidered the same as viciae Kaltenbach. It certainly is very similar in every respect. Viciae Kalt. was set as the type of liis genus Drepaniella by Del Guercio, therefore Drepaniella will become a synonym of Megoura. Characters. — Head with distinct antennal tubercles. Antennae of six segments, armed with subcircular sensoria. Wing venation normal; cornicles moderately long and swollen in the middle. Cauda about the same length as the cornicles. Type (nionotypical), Megoura viciae Buckton. Genus MYZUS Passerini. Plate Vm, L, M, X-Z. 18G0. Myzm Passerini, Gli Afldi, p. 27. 1800. Rhopalosiphum Passerini, Gli Afidi, p. 27. 1913. Myzoidcs Van der Goot, Tijd. voor Ent., v. 50, p. 81. 1913. Ovatus Van der Goot, Tijd. voor Ent., v. 56, p. 84. 1011. i/yzofZc.s Mordwilko, Faune Buss. Aphidoidea, p. 52. 1914. Aulacorlhum Mordwiiko, Fauna Russ. Aphidoidea, p. 5S. 1910. Ncomyzus Van der Goot, Zur Kennt. der Blattlause Java's, p. 50. 1918. Myzopsis Matsumura, Trans. Sapporo Nat. Hist. Soc, v. 7, pt. 1, p. 19. 58 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE, The genus Myzus was erected with ApMs cerasi Fab, as type. In the same year Passerini used Aphis ix'rsicae Sulz. as type of Rhopalosiphum Koch. Ajyliis nympliaeaelj., however, had been set as the type of that genus m 1856. Persicae seems to be very closely related to cerasi and Rhopalosiphum (Koch) Pass., therefore, is placed as a synonym of Myzus. In 1913 Van der Goot placed certain other species in Myzus and the type of the genus he made the type of his Myzoides. This will evidently then become a synonym, as it has the same type. In 1916 Van der Goot erected the genus Neomyzus with circum- jlexum Buckt. as type. While the frontal tubercles of this species do not converge quite to the same extent as those of cerasi, they are' quite similar and the cauda is somewhat conical as in that species. Neomyzus, therefore, is believed to be a synonym. Mordwilko used the name Myzodes with tahaci Mord. as type. This species he does not fully describe, but gives a figure and describes the characters of the genus. From the information given it appears to be a synonym of Myzus. The writer also believes that the genus Ovatus V. d. Goot is a synonym of Myzus. This genus was erected with mespili V. d. Goot as type. The genus Aulacorthum Mord. was erected with pelargonii as type. A study of this species shows the antennal tubercles very similar to those of cerasi. The cornicles, too, are quite similar, although the cauda is a little more Aphis-like. The writer believes this genus is a synonym. Characters. — Head with distinct antennal tubercles present which, particularly in the apterous form, project inward and are strongly gibbous. Antennte of six seg- ments, the first segment gibbous like the antennal tubercles. Wing venation normal. Cornicles rather long and subcylindrical. Cauda somewhat short and conical, con- stricted very slightly, if at all. Type (fixed by Passerini, 1860), Aphis cerasi Fab. Genus PHORODON Passerini. Plate VIII, N-Q. 1800. Phorodon Passerini, Gli Afldi, p. 27. Characters. — Head in the alate form with distinct antennal tubercles which project somewhat inward, first antennal segment gibbous. In the apterous form the antennal tubercles possess very prominent projections which extend forward in front of the head. First antennal segment with a projecting process. Fore wings with media twice branched; hind wings with both media and cubitus present. Cornicles cylin- drical, in the apterous form somewhat curved; cauda rather acutely conical, not as long as the cornicles. Males as a rule winged and oviparous females apterous. Type (fixed by Passerini, 1860), Aphis humuli Schrk. Genus RHOPALOSIPHONINUS, n. gen. Plate IV, D-F. The genus Rhopalosiphoninus is erected for latysipJion Davidson, a species with very peculiar cornicles. It appears to be somewhat related to Amphorophora. GENERIC CLASSIFICATION OF APHIDIDAE. 59 Characters. — Head with promineut antennal tubercles which project inward and are armed with prominent spines. Antennse of six segments armed with subcircular sensoria and in the first segment with spines similar to those of the antennal tubercles. Wing venation normal. Cornicles narrow and cylindrical at the base, then abruptly and prominently swollen, returning again abruptly to the normal size near the tip. Cauda rather short and conical. Type, Amphorophora latysiphon Davidson. Subtribe PENTALONINA. The subtribe Pentalonina is one of the Aphidini in which speciahzation in the wing venation has taken place in a pecuhar manner. The radial sector has in one genus extended downward and coalesced with the upper branch of the media. In the genus Idiopterus the two have not become entirely fused, though in some specimens they have almost done so. In Pentalonia, however, the veins have become permanently united, and a very peculiar-looking venation is the result. A closed cell is formed by the radial sector when it meets the upper branch of the media, and when it leaves this again it gives a three-branched appearance to the upper branch of the media. The explanation of this peculiar venation is, however, easily understood by comparison with the venation of Idiopterus. In some of the genera the hind wings are greatly reduced, so that some- times only one vein remains, while in other genera this reduction has not taken place. Most of the wing veins are clouded with brownish borders. The insects feed usually upon ferns or tropical plants. Key to the Genera of the Pentalonina. 1. Hind wings much reduced in size, lacking at least the cubitus 2. Hind wings nearly normal in size and with both media and cubitus present. 3. 2. Radial sector of fore wing fused with the upper branch of the media, forming a closed cell Pentalonia,. Radial sector of fore wings not so fused, but normal Microparsus. 3. Cornicles cylindrical Idiopterus. Cornicles somewhat swollen near their distal extremities 4. 4. Media of fore wings twice branched Fullawayella. Media of fore wings once branched Neotoxoptera. Genus FULLAWAYELLA Del Guercio. 1911. Fullawayella Del Guercio, Redia, v. 7, p. 462. 1916. Micromyzus, Van der Goot, Zur Kenntniss der Blattlaiise Java's, p. 55. This genus is very suggestive of Amphorophora in certain ways but no doubt is related here. Van der Goot's genus was erected with nigrum V. d. Goot as type but this species differs very little from hirkaldyi. Characters. — Antenna? on prominent, converging, imbricated antennal tubercles, of six segments and armed with subcircular sensoria. Fore wings with the media twice branched and with the radical sector deeply curved toward the upper branch of the 60 BULLETIN 826^ U. S. DEPARTMENT OF AGEICULTURE. media ; hind wings with both media and cubitus present. Cornicles somewhat swollen near their distal extremities. Cauda elongate and constricted near the base. Type (monotypical), 3f acrosiphum hirlcaldyi Tullav^ay. Genus IDIOPTERUS Davis. Plate VIII, EE-HH. 1900. Idiojncru^ Davis, Ann. Ent. Soc. Amer., v. 2, p. 19S. IJiopterus, a less specialized genus tlian Pentalonia, is worthy of special notC; as it gives a key to the peculiar venation of the latter genus. The coalescing of the radial sector and the media is here plainly visible and in some specimens a triangular closed cell is formed, although in most examples the two veins can he traced distincth^. Characters. — Head ^vith prominent antennal tubercles which project slightly in- wards, and are gibbous. Antennae of six segments, armed with subcircular sensoria, the first segment gibbous like the antennal tubercles. Cornicles subcylindric, rather slender, cauda somewhat elongate, conical. Fore wings with the radial sector extend- ing abruptly downward from the stigma and paralleling the upper branch of the media Avith which in some specimens it appears to be almost united ; hind wings with both media and cubitus present. Tj'pe (monotypical), Idiopterus nephrolepidis DaAds.. Genus MICROPARSUS Patch. Plate VIII, AA-DD. 1909. Microparsus Patch, Ent. News, v. 20, p. 337. !Microparsus is at once distinguished from the other genera related to it hy the peculiar venation and the reduction of the hind ^dng. Characlers. — Head with distinct antennal tubercles present. Antennte of six seg- ments, armed with subcircular sensoria. Fore wings with the media once branched; hind wings much reduced in size and lacking both the media and cubitus. Cornicles subcylindric. Cauda rather long and tapering, almost equal in length to the cornicles. TA'pe (monotypical), Microparsus variabilis Patch. Genus NEOTOXOPTERA Theobald.i 1915. Neotoxoptcra Theobald, Bui. Ent. Res., v. 6, p. 131. This genus is closely related to FuUawayella, from which it can be separated by the venation. There has been some doubt whether or not this is a good genus, for the name wiH not hold if the type is found to correspond with Pergande's violae, which it resembles. In that case Neotoxoptera would become a synonym of FuUawayella, for Pergande's species is undoubtedly a FuUawayella. Characters. — Head with prominent antennal tubercles. Antennae of six segments, armed with subcircular sensoria. Fore wings with the media once l)ranched. Corni- cle? swollen, elongate, somewhat conical. Type (monotypical), Neotoxoptera violae Theo. 1 After thi.s paper was in type the writer (Bui. Ent. Res., v. 10, p. 45) showed that violae Theo. is a Sjrnonym of violae Perg. GENERIC CLASSIFICATION OF APHIDIDAB. 61 Genus PENTALONIA Coquerel. Plate VIII, II-MM. 1859. Pentalonta Coquerel, Arm. Ent. Soc. France, Ser. 3, v. 7, p. 259. The genus Pentalonia Coquerel is a very peculiar one and possesses a venation unlike that of a,nj other in the Aphididae. It is, however, only a little further development of the condition met with in Idiopterus, which is the less specialized of the two genera. Characters. — Head with prominent antenna! tubercles which are, more especially in the apterous form, projected inward, gibbous and somewhat Myzus-like in appear- ance. Antennje of six segments, armed with subcLrcular sensoria, the first segment gibbous like the antennal tubercles. Cornicles somewhat constricted near their middle, then again somewhat swollen near their distal extremity. Cauda rather small but elongate, subcorneal, slightly constricted about the middle. Fore wings with the radial sector extending abrtiptly downward and meeting theupper branch of the media with which it fuses but is diverted again toward its natural course near the tip of the wing. A closed cell is thus formed by the radial sector and the media but at the margin of the wing there are the same veins as in the Aphidini (Plate VIII, JJ.) Hind wings very much reduced, cubitus absent. T\'pe (monotypical), Pentalonia nigronervosa Cql, Subfamily II, MINDARINAE. It has been the custom to consider the genus Mindarus as closely re- lated to the Pemphigini, but the writer is unable to do this and concludes that it must represent a subfamily in itself. In some waj^s abietinus is the most primitive livmg aphid. It is, in fact, the only one which has retained the general wing structure which is predominant in the fossil forms. It is true that the venation is more reduced than in some of the other subfamihes, but the type of wmg in regard to the stigma formation is exactly like most fossil wings and unlike the wings of other living forms. Many of the characters suggest the Eriosomatinae and the genus is no doubt very similar to the ancestors of the insects in that subfamily. The antennal structure and general form are like those in the Eriosomatinae. The sexes, too, are apterous, but though they have developed the small apterous condition they are in many ways more primitive than are the sexes of the Eriosomatmae. The male is small and suggests the condition in those forms. The peculiar habit of copulation is similar, in that the male mounts the female and may remain there inactive for a very long period. The writer has observed a male of Eriosoma lanigerum clinging thus to a female for 48 hours. The sexes of Mindarus, however, have not lost the beak and the male feeds on the juices of its host. In this regard they are more primitive than sexes in the Eriosomatinae. The oviparous female, moreover, develops her ovaries and produces as high as 8 or 9 eggs, in striking contrast with the ovipara in the Eriosomatinae. It is a much less specialized condition. In regard to the alate form the shape of the cauda is quite different from that met with in the Eriosomatinae. 62 BULLETIN 826^ IT, g. DEPARTMENT OF AGRICULTURE. It seems to the writer that the Mmdarmae give a fair idea of tlie ancestors of the Eriosomatinae and may even represent a group dominant in earlier times from which the Eriosomatinae sprang. Only one genus is represented. Genus MINDARUS Koch. Plate IX, A-F. 1857. Koch, Die Pflanzenlause Aphiden, p. 277. The peculiar genus Mindarus was erected by Koch with ahietinus Koch as type. This species is the only one in the genus, although it has been redescribed as ScTiizoneura pinicola Thos. and Schizoneura obliqua Choi. Characters. — Cornicles present as mere rings. Large wax plates present. Alate forms with six-segmented antennae armed with oval sensoria. Fore winge with the media once branched; radial sector inserted mesad of the long narrow stigma, thus giving a very long stigmal cell; hind wings with both media and cubitus present. Cauda rather long, not rounded, but somewhat conical or even spatulate. Sexes small and apterous, beaks present and feeding taking place. 0-viparous female with the ovaries developed and lading as high as 9 eggs. Forms living free upon the twigs of conifers which become somewhat distorted by the feeding of the insects. Tj'pe (monotypical), Mindarus ahietinus Koch. Subfamily III, ERIOSOMATINAE. The subfamily Eriosomatinae is com])osed of insects which are perhaps as specialized as any of the Aphididae. Tliey show a re- markable development of the habit of gall formation and in this respect parallel the Hormaphidinae. The insects of that subfamily, however, evidently have developed the habit independently. Many previous authors have placed all of these forms in the present sub- family. This, the writer beUeves, is incorrect, as shown by the biol- ogies of the insects. The sexual forms give a true understanding of the relationships and of the genera w^hich should be included in the Eriosomatinae. All of the forms included by the writer show evi- dence of a common origin in that the sexes have become degenerate. They have become small apterous forms and have lost the mouth parts and the ability to take food. That this was not their original condition is clearly shown by the history of the family and also by the fact that the sexual forms of some species have a beak when born, but lose this at the first molting. Other species even at the time of birth are devoid of all but a rudimentary trophictubercle. The reproductive system of the female has become greatly altered. As previously pointed out by the wi'iter, the early development of the reproductive system of the sexual female corresponds exactly to that in the apterous forms and to that of the oviparous forms of the more primitive groups. Young embryos * * * show that the ovaries are at first similar to those of the parthenogenetic form. There may he distinguished the four chambers on each side containing egg cells and nutritive cells. In later embryos most of the egg tubes are in GENERIC CLASSIFICATION OF APHIDIDAE. 63 the process of degeneration and only two ovaries, one on each side, develop. Of these one finally degenerates and the egg of the other grows until it fills almost the entire body and the insect appears to be little else than egg.''^ It will be seen at a glance that such a method of egg development is entii-ely different from that met with in members of the genera which have been heretofore placed in the subfamily. The Hormaphidini and the Thelaxini, as will be seen under the discussion of those tribes, have sexual females which develop normal ovaries and lay several eggs in the same way as do the Aphidini, Lachnini, and other groups. It is true that some have developed gall formation and highl}^ special- ized, wax-secreting organs, but this is more of a parallelism than a close relationship, as is indicated by the sexual forms. The wax- secreting organs of the Eriosomatinae vary considerably in structure. A study of those in the genus Eriosoma has been presented by the writer (1915). The glands here are compound, each cell containing a central wax chamber into which the wax is secreted and from which it is forced out as a fine waxen thread. In other genera the wax glands take on the nature of plates, illustrated in the genus Prociphilus. These glands are essentially the same in general struc- ture as are those in Eriosoma, but the wax cells are placed very close together and are so extremely elongate that their openings to the surface are very small. A large number of these gives the appearance of a more or less uniform plate. The structure, however, in the two genera follows the same lines. The wing venation in tliis subfamily presents as great a reduction as in any of the subfamilies of the Aphididae and in this respect it is comparable to the Hormaphidinae. In the fore wings the reduction is shown in the media which is never branched more than once. Dr. Patch has pointed out the homologies of the veins and has indicated that in all of these cases the branches represent M1+2 and M3+4. In some cases, however, it would appear as if they were M^ and M4. In other genera the media is indicated as a single vein. The radial sector is in nearly every case present and the cubitus and first anal are prominent veins. The tracheae are figured for the subfamily under the genus Eriosoma. In the hind wings the radial sector is always present and two oblique veins are nearly always found. These are the media and the cubitus. In several genera, however, the cubitus has disappeared and only the media remains as the one transverse vein in the hind wings. The cornicles in the genera of this subfamily are not j^rominently developed. Indeed, they are absent altogether in certain of the tribes. In the genus Eriosoma they are chitinized rings slightly elevated on shallow hauy cones. The opening of the cornicles is closed by a muscle and from the cornicle a narrow duct leads to a > Baker, A. C. The woolly apple aphis. U. S. Dept. Agr., Off. of Sec, Kept. 101, p. 43. 1915. 64 BULLETIlSr 826^ U. S. DEPARTMENT OF AGRICULTURE. large wax reservoir. The structure of the cornicles themselves in this subfamily is essentially the same in all genera where they are present. In a large number of genera, however, the wax reservoir is absent and in some specialized tribes the cornicles are likewise absent. It is interesting to note that in some genera, though absent in the stem mother, they are present in the alate forms. The habit of gall formation is not found equally in all genera and it would seem that those forms which have become associated with arits have not developed this habit to the same extent as have some of the other groups. However, it must be borne in mind that our knowledge concerning the species associated with ants is very incom- plete, and the writer is convinced that many of the Prociphilini will be found during their summer generations in this relation. Many of the forms cause true galls which are the result of outgrowths of the plant and which completely enclose the insects. Sometimes the stem mother lives in a gall by herself while in other cases the following generations live with her. The original spring gall is usually the result of the activities of the young stem mother. Certain species do not produce true galls but form pseudogalls which are due to the rolling or crumpling of the leaves on which the insects feed. Other species, again, especially during their summer generations, feed on the twigs or roots of plants and give rise to excrescences by their feeding. It often happens that species which in their spring forms are gall makers, attack plants in this way in their summer generations. Others live on the roots of grasses during these generations and do not cause the excrescences produced by those species feeding on trees and woody shrubs. The association with ants is highly developed by one tribe of this subfamily, although all of the other tribes are to a degree tended by these insects. The species of the Fordini live exclusively in the nests of ants or are tended by them, and they are cared for very carefully in return for the honeydew excreted. Ants also attend species which have aerial feeding habits and they may be seen carrying the root generations of species of Eriosoma from one place to another and even distributing them about on the trees. Indeed the wi-iter once took advantage of the presence of ants to infest some apple seed- lings. A vial of apterous insects was emptied at the base of each tree and the ants soon could be seen running about carrying the apliids to suitable positions on the trees. Sometimes, however, they carried them away. GENERIC CLASSIFICATION OF APHIDIDAE. 65 Key to the Tribes op the Eriosomatinae. 1. Cornicles present, at least in the alate forms; however, often mere rings 4. Cornicles absent 2. 2. Forms liWng in true galls or in pseudogalls on plants 3_ Forms living in the nests of ants or at least subterranean, feeding on the roots of plants; wax-secreting areas present; antennae of alate forms rather short and thick with somewhat oval sensoria Fordini. 3. Forms li^ang in true galls and without wax plates prominently developed on the head and thorax of alate form; wax-secreting areas present but not prominently developed. Alate forms lea\dng the galls in the late summer or fall. Antennae with annular sensoria Melaphini. Forms living in pseudogalls, occasionally in true galls. Antennae of alate form rather long and slender, with narrow or somewhat oval or rounded sensoria . Wax plates well developed and present on the head and thorax of the alate forms which leave the galls in the spring Prociphilini. 4. Forms li\'ing in galls, pseudogalls, or free upon their host; wax glands prom- inently developed; antennae of alate forms armed with annular sensoria which almost completely encircle the segments Eriosomatini. Forms living usually in true galls; wax glands present but not strongly de- veloped; antennae of alate forms armed with narrow, transverse sensoria, somewhat oA-al or irregular ones, or occasionally without sensoria. .Pemphigini. Tribe ERIOSOMATINI. The tribe Eriosomatini is composed of insects which have more or less developed the habit of gall formation, which are possessed of wax glands, and the antennae of the alate forms of which are armed usually with annular sensoria. Their typical host group is that of the elms. Characters. — Forms living in galls, pseudogalls, or free upon the twigs or roots of their host on which they form excrescences. Prominent wax glands present. Cor- nicles distinct; antennae of alate forms armed with annular sensoria which of ten almost completely encircle the segments. Sexual forms small, apterous, beakless; oviparous females developing a solitary egg. Key to the Genera of the Eriosomatixi. 1 . Media of the fore wings of alate form once branched 3. Media of the alate form simple 2. 2. Hind wings with both media and cubitus present Gobaishia. Hind wings -with only the media present Tetraneura.^ 3. Hind wings with both media and cubitus present Eriosoma. Hind wings with only the media present 4. 4. Stem mother with four-segmented antennae; antennae of alate form rather short and thick Colopha. Stem mother with five-segmented antennae ; antennae of alate form long and slender '. Georgia. 1 There is considerable evidence for separating a tribe Tetraneurini to include the genera Colopha, Tetraneura, and Gobaishia. 141G13°— 20— Bull. S2G — —5 66 BULLETIN 826; U. S. DEPARTMENT OF AGRICULTURE. Genus COLOPHA Moaell. Plate IX, G-L. 1877. Colopha Monell, Can. Ent., v. 9, p. 102. The genus Colopha was erected for ulmicola Fitch. One of the principal characters whereby it may be separated from Tetranem^a is the once-branched character of the media. The two genera, how- ever, are very closely related. The species have the same peculiar structure and the same mode of life. Characters. — Cornicles slightly elevated rings. Stem mother with four-segmented antennae. Apterous form with five-segmented or sometimes six-segmented antennae. Wax glands present. Alate form with six-segmented antenna? which are armed with annular sensoria partly encircling the segments. Fore wings with the media once branched, hind wings with only the media present. Forms making galls upon the leaves of trees in which the stem mother and her offspring live in company ; in summer migrating to the roots of plants. Type (monotypical), Byrsocrypla ulmicola Fitch. Genus ERIOSOMA Leach. Plate IX, M-T. 1S18. Eriosoma Leach, Trans. Hort. Soc. London, v. 3, p. 60. 1831. Myzoxylus Blot, Mem. Soc. Roy. Agr. ct de Com. Caen., v. 3, p. 332. 1837. Schiznncum Hartig, Jahresb. ii. d. Fortschr. d. Forstwiss. und forstl. Naturk., v. 1, p. 645. 1848. Mima-phidus Rondani, Nuovl Annali della Seienze Natm'ali, ser. 2, v. 9, p. 35. In 1818 Leach erected his genus in a footnote in connection with a paper read by Mosley. The paper was published in 1818. In 1819 Samouelle published his ''Useful Compendium" and on page 232 characterized the genus Eriosoma Leach MSS. The prmted copy of the Transactions appeared complete in 1820. In 1824 Blot used the word Myzoxyle which he corrected to Myzoxylus in 1831. For these Aphis lanigera Hausm. was used as tyi^e. In 1837 Hartig erected Schizoneura and of this genus uhni was made type by Passerini in 1860. This then wiU become a synonym. Corni Fab. was for a time considered the type of this genus but this species was not in the original genus. In 1848 Rondani used Mima- phidus with ulmi Fab. as type, which according to Passerini is the same as lanuginosa Hartig. Therefore, this genus will become a synonym. Characters. — Comicles distinct rings on somewhat elevated tubercles. Apterous form with six-segmented antennae. Stem mother with five-segmented antennae. Wax plates present in the apterous and alate Ad\apara. Alate form with six-seg- mented antennae armed with annular sensoria. Fore wings with the media once branched, hind wings with both media and cubitus present; cauda and anal plate rounded. Forms li\dng in gall-like formations or causing excrescences on their hosts. Sexual forms small, apterous, beakless. Only one egg of those of the oviparous female develops. Type (monotypical), Aphis lanigera Hausmann. GENERIC CLASSIFICATION OF APHIDIDAE. ^7 Genus GEORGIA Wilson. Plate IX, U-Z. 1911. Georgia Wilson, Can. Ent., v. 43, p. 64. The genus Georgia appears to be related to Colopha Mon., and yet many of the characters are so like those of Eriosoma that the insect suggests that genus also. Especially'' to the species E. amerlcanum Riley there is a striking resemblance. Prominent wax glands are lacking but these are sometimes also lacking in the spring forms of Eriosoma. Characters. — Cornicles present and situated on shallow hairy cones as are those of Eriosoma. Stem mother with five-segmented antenn&e ; alate form with six-segmented a'ntennae which are armed with narrow sensoria that do not encircle the segment to any extent. Fore wings with the media once branched, hind wings with only one oblique vein. Cauda rounded. Prominent wax pores such as those present in Eriosoma lacking, but small wax areas present. Forms living in pseudogalls on plants, the alate individuals migrating from the galls in the early spring. Tj'i^e (monotj^ical), Georgia ulvii ^Isn. Genus GOBAISHIA Matsumura. Plate X, A-G; XI, V. 1909. Byrsocrypta Tullgren, Arkiv for Zoologi, Bd. 5, no. 14, p. 182. 1917. GobaisMa Matsumura, Synopsis of the Pemphigidae of Japan, Gifu, Japan, p. 75. Tullgren used the name Byrsocrypta Hal. as the name of a sub- genus with pallida as type, placing it under Tetraneura. He appar- ently overlooked the fact that Westwood had set hursarius as the type of the genus Byrsocrypta as will be found discussed under the genus Pemphigus. Pallida is different from the species of Tetraneura, in that the cubitus is retained in the hind wing. If Colopha is retained on account of the branched nature of the media in the fore wings it will be necessary to place pallida as typical of a genus related to Tetraneura. In 1917 Matsumura erected the genus Gobaishia with Gohaishia japonica Mats, as type. This species was stated to be very similar to Tetraneura alba Ratz. Tetraneura alba Ratz is the same species as Eriosoma pallida Haliday and the characters given for the genus are, therefore, similar to Tullgren 's conception of Byrsocrypta. The figures drawn are from specimens of pallida as no japonica was avail- able to the writer for study. Characters. — Cornicles present, stem mother with four-segmented antennae, alate forms with six-segmented antennae which are armed with annular sensoria. Fore wings with the media usually simple; hind wings with both cubitus and media present. Type (fixed by Matsumura, 1917), Gobaishia japonica Mats. 68 BULLETIN 826; XJ. S. DEPARTMENT OF AGRICULTURE. Genus TETRANEURA Hartig. Plate X, H-M. 1841. Tclraneura Hartig, Germar's Zeitschrift fiir die Entomologie, v. 3, p. 366. In 1841 Hartig erected tlie genus Tetraneura under which he gave Tetraneura ulmi Lin. ? questioned thus and described. He also hsted T. rugicornis Hartig. One of these species was questioned and the other merely listed. ZTlmi L., however, was questioned only in the sense of the determmation, and a good description was given so that it is known what insect Hartig had. In 1843 Kaltenbach gave a description of the genus Tetraneura crediting it to Hartig and described thereunder one species, Aphis ulmi De Geer. Aphis ulmi De Geer (1773) is the same species as Aphis ulmi Geoff roy (1764) but this name can not be used, smce Linnaeus used Apliis ulmi for a different insect. This is the same insect de- scribed by Hartig as T. ulmi L. ? and it is evident that it requires a new name, to which ulmifoliae is given. Characters. — Cornicles very slightly elevated rings, not at all prominent. Stem mother with four-segmented antennse; apterous form with five-segmented antennae. Wax glands present. Alate form with six-segmented antennae wliich are armed with narrow annular sensoria almost completely encircling the segment. Fore wings with the media simple; hind wings with only the media present. Forms living in galls and migrating in spring to other plants. Sexes small, apterous and beakless. Oviparous female developing only one egg. Type, Tetraneura ulmifolix Baker {Aphis ulmi L. of Hartig). Tribe PEMPHIGINI. The tribe Pemphigini is composed of forms which are highly spe- cialized and most of which have developed the habit of true gall formation. The secretion of wax also occurs but wax secreting plates are not developed to the extent met with in some of the other tribes of the subfamily. Alternation of hosts is found to occur, migrants leaving the galls in early sprmg or summer and returning in autumn. In some species, however, the insects do not leave the gaUs until the mothers of the sexual forms are produced. Distinct cornicles are present and by this character forms in some of the other tribes which are suggestive of the Pemphigini may be distinguished. The typical host group is Populus and the galls are normally spring galls. Characters. — Forms usually inhabiting true galls and often migrating to other plants during the summer. Antennae of six segments in the alate form and in nearly all genera armed with linear, oval, or somewhat irregularly shaped sensoria. Small wax- secreting areas present. Sexual forms small, apterous, and beakless, the o\'iparous female developing only one egg. Six genera may be included in the tribe and these genera may be separated by the following key : GENERIC CLASSIFICATION OF APHIDIDAE. 69 Key to the Genera of the Pemphigini. 1. Unguis of segment VI of alate form distinctly long and Aplus-like. .Mordwilkoja. Unguis sliort and knob-like 2. 2. Media once branched 3. Media simple 4. 3. Antennae of alate form usually without secondary sensoria. Wings flat in repose Phloeomyzus. AntenuEe of alate form with secondary sensoria. Wings not flat in repose Pachypappella. 4. Both media and cubitus present in hind wing 5. One oblique vein only in hind wing Dryopeia. 5. Antennae of alate form without secondary sensoria. Wings flat in repose Rhizoc tonus'. Antennise of alate form with secondary sensoria. Wings not flat in repose ... 6 . 6. Antennae of alate form rather short and thick. Stem mother with four-seg- mented antennae .• Pemphigus. Antennae of alate form rather long and slender. Stem mother with five-seg- mented antennte Cornaphis. Genus CORNAPHIS Gillette. Plate X, N-T. 1913. Cornaphis Gillette, Ann. Knt. Soe. Am., v. 6, p. 491. The genus Cornaphis was erected by Gillette for his species Corna- phis popidi. In his tlescrii^tion it is stated that the genus is closely related to Asiphum. In Cornaphis, however, there are large corni- cles in the alate form and in other respects it seems that the genus is closely related to Pachypappella. In that genus, however, the media is once forked, whereas in Cornaphis it appears to be simple, at least as a rule. This difference has led the wi-iter to retain a genus with lactea as type rather than to place that species and similar ones in Cornaphis. Characters. — Cornicles present; stem mother with five-segmented antennae and without wax plates. Alate form with six-segmented antennae armed with rather narrow sensoria; permanent sensoria ciliate. Fore wings with the media simple, hind wings with both media and cubitus present; wax plates present in the apterous form; sexes small, aj^terous and beakless; the oviparous female developing only one egg. Forms living in galls, the stem mother and the following forms living in the same gall. Tyi^e (monotypical), Corna2Jhis populi Gill. Genus DRYOPEIA Kirkaldy. Plate X, U-Y. 1857. Endeis Koch, Die Pflanzenlause Apliiden, p. 312. 1889. Endeis Ashmcad, Ent. Amcr., v. 5, p. 189. 1901. Dryopeia Kirkaldy, The Entomologist, v. 37, p. 279. 1917. Watahura Matsumui'a, Synopsis of the Pemphigidae of Japan, p. 89. In 1857 Koch erected his genus Endeis with two species, heUa Koch and rorea Koch. This name was replaced by Dryopeia in 1904 by Kirkaldy, and hella has been definitcl}^ l^laced as the type. 70 BULLETIN 826^ U. S. DEPARTMENT OF AGRICULTURE. In some respects the genus is suggestive of Anoecia, altliough it seems to be undoubtedly a Pemphiginid and will no doubt be so proven by the sexual forms. In 1917 Matsumura erected his genus Watabura with Watahura nishiyae Mats, as type. This species was stated to resemble a Pem- phigus, excejDting that the antennal segments are somewhat different and only one oblique vein is in the hind wing. (Two obliques are shown in his PI. XII, 9). The antennae are armed with narrow trans- verse sensoria and there seems little doubt that this genus is a syno- nym of Dryopeia. It is noteworthy that the life history of the type species is not known, but it is thought to live on the roots of trees. The type of the genus Dryopeia is a root feeder. Characters. — Cornicles present, situated on broad shallow cones, suggestive of those of Anoecia. Stem mother unknown, apterous forms with six-segmented antennse. Alate forms with six-segmented antennse, armed with narrow transverse sensoria. Fore wings with media simple, hind wings with one oblique vein. Summer forms subterranean, living on the roots of plants. Spring forms and sexes unknown. Apter- ous tarsi one-segmented. Type (fixed by Kirkaldy. 1906), Endeis hella Koch. Genus MORDWILKOJA Del Guercio. Plate XI, A-G. 1909. ITordivilkoja Del Guercio, Rivista Patol. Veget., v. 4, p. 11. This genus was erected in 1909 for the peculiar species Byrsocrypta vagahunda Walsh. This differs in the antennjB quite remarkably from all of the other si3ecies belonging to this tribe. The difference is in the long unguis of the sixth segment. However, the other char- acters and the four-segmented nature of the antennae of the stem mother seem to place it with little doubt in the Pemphigini. There has been some doubt cast by Oestlund on the determination of Walsh's species and this has led Cockerell to propose the name oestlundi for the species now known so well, but, as Gillette has pointed out, Walsh evidently accepted the insect of Riley and Monell as the same species as his vagahunda: The insects Riley had were undoubtedly the species we know and the wiiter therefore accepts vagahunda and the generic name Mordwilkoja. The genus was erected with the name vagahunda used as type and not oestlundi. Characters. — Cornicles present as somewhat elevated rings. Stem mother with four- segmented antennae, the unguis of segment VI slender and Apliis-like. Permanent sensoria ciliate. Alate form with five-segmented antennse which are armed with narrow transverse sensoria. Fore ^Ying^ with the media simple, hind wings with both media and cubitus present. Forms li^ing in galls; the stem mother and her offspring living in the same gall, the alate forms leaving the galls in spring or early summer. Sexes unknown, but no doubt small, apterous, and beakless. Type (raonotypical), Byrsocrypta vagahimda Walsh. GENERIC CLASSIFICATION OF APHIDIDAB. 71 Genus PACHYPAPPELLA, n. n. Plate XI, H-M. 1909. Pachypappa Tullgren, Arkiv for Zoologi, Bd. 5, no. 14, p. 69. In 1854 the genus Pachypappa was erected by Koch with marsu- pialis and vesicalis in the genus. A study of marsupialis shows that this species is in reality a Pemphigus as it shows all the characters of this genus. Tullgren, 1909, noted this and therefore interpreted the genus differently. Marsupialis had, however, been set as the type of the genus. Pachjq^appa Koch, therefore, becomes a syno- nym of Pemphigus, and Pachypappa Tullgren must receive a new name for which Pachypappella is here given. Characters. — Stem mother without cornicles but with wax plates; antennae five- segmented. Alate form with cornicles; antennae six-segmented and with transverse sensoria. Fore wings with media once branched, hind wings with both media and cubitus present. Type (present designation), Pachypappa lactea Tullgren. Genus PEMPHIGUS Hartig. Plate XI, N-U. 1837. Pemphigus Hartig, Jahresb. u. d. Fortschr. d. Porstwiss. und forstl. Xaturk., v. 1, p. 645. 1839. Byrsocrypta Haliday, Ann. Nat. Hist., v. 2, p. 190. 1840. Brysocrypta Westwood, Int. Mod. Class. Ins., Synopsis, v. 2, p. 118. 1847. Aphioidcs Rondani, Nuovi Annali Sci. Nat. Bologna (2), v. 8, p. 439. 1857. Amycla Koch, Die Pflanz. Aphiden, p. 301. 1857. Pachypappa Koch, Die Pflanz. Aphiden, p. 263. 1857. Rhizomaria Hartig, Verhandl. d. Hils-SoUing-Forstvereins, Jahrg. 1856, p. 52. 1859. Tychea Koch, Die Pflanzenlaiise Aphiden, p. 296. 1885. Kessleria Lichtenstein, Mon. Puceron du Peupl., p. 16. 1904. Hamadryaphis Kirkaldy, The Entomologist, v. 37, p. 279. In 1837 Hartig erected his genus Pemphigus, although it was not until 1841 that his reference to the genus as generally cited appeared. Passerini in 1860 set hursarius as type. In 1839 Haliday used the generic term Byrsocrypta but mentioned no species. In 1840 West- wood referred to this genus as Brysocrypta and gave hursaria L. as type. In 1859 Koch erected the genus Tychea with graminis Koch as type (mono typical). Schouteden (1906) has described the winged form of Tychea graminis Koch and stated that it is a typical Byrsocrypta. The writer has had no opportunity to study specimens but on the strength of tliis statement of Schouteden places Tychea as a synonym of Pempliigus. It is worthy of note, however, that Schouteden did not mention the cornicles, and this is a point of considerable difference if graminis is a Pempliigus or if it belongs to the Fordini. In 1857 Hartig described the genus Rhizomaria with inceae Hartig as type. This species, however, appears to be a typical Pemphigus and Rhizomaria will become a synonym. In 1857 Koch erected the genus Amycla and of tliis gewu^ fuscifrons Koch has been made the type. The writer has been unable to obtain specimens of this species but from the descriptions it seems almost 72 BULLETIN 826;, U. S. DEPARTMENT OF AGRICULTURE. certain that this species is a true Pemphigus. This will thus make the genus Amycla a synonym. In 1847 Rondani described the genus Aphioides of which bursaria Fab. was indicated as the type and Aphioides, therefore, is a synonym. In 1854 Koch erected the genus Pachypappa of wliich marsuinalis Koch has been made the type. Marsupialis, however, is a typical Pempliigus. Pachypappa Koch, therefore, must become a synonym. Tullgren (1909) used Pachypappa in a different sense, but this is discussed under the genus Pach3^pappella. In 1886 Lichtenstein erected the genus Kessleria for spirotliica and this name was replaced by Hamadryaphis Kirk, in 1904. A study of this species, however, shows that it is a typical Pemphigus. Therefore, these two names will become synonyms. Characiers. — ^Cornicles present; wax plates, if present, weakly developed; stem- mother with four-segmented antennae ; alate form with six-segmented antennae which are armed with narrow, oval or somewhat irregular sensoria. Fore wings with the media simple; hind wings with both media and cubitus present Sexes small, apter- ous, and beakless. Oviparous female developing only one egg. Forms living in galls, the stem-mother and her offspring in the same gall, the alate forms typically leaving the galls in the spring. Type (fixed by Passerini, 1860), Aphis bursaria L. Genus PHLOEOMYZUS Horvath. Plate XI, W-BB. 1886. Lowia Lichtenstein, Mon. Puceron PeupL, p. 37. 1S96. Phlocomyzus Ilorvath, Wien. Ent. Zeit., v. 15, p. 5. In 1886 Lichtenstein erected the genus Lowia with Schizoneura 2)asserinii Sig. as type but as tliis name had been used previously it was replaced in 1896 by Phloeomyzus Ilorvath. It is with some hesitation that the writer places this genus in the Pemphigini. In some respects it suggests the Melaphini, wliile in many respects it strongly suggests the Thelaxini or even the Phyllapliidina. Indeed, to the Melaphini it shows striking resemblances. Without a study of the sexual forms it will be very difficult to place the genus definitely. All that can be done at the present time is to place it tentatively with the forms with which it appears to be related, and if further study shows this to be incorrect the genus can be placed definitely with its allies. Characters. — Cornicles present, very slightly elevated. Apterous form with six- segmented antennfo. Alate form with six-segmented antennse which are rather slender and without secondary sensoria. Fore wings with the media once branched, hind wings with both media and cubitus present. Large wax plates present on the abdomen. Wings held flat in repose. Forms living free upon the bark of trees in colonies. Type (monotypical), Schizoneura 2)asscri)iii Sig. Study based on specimens received from Mordwilko from Warsaw, Poland, and notes by Pergande on type specimens loaned by Horvath. GENERIC CLASSIFICATION OF APHIDIDAE. 73 Genus RHIZOCTONUS Mokrzecky. 1895. lihizoctonus Mokrzecky, Horae. Soc. Ent. Ross., v. 30, p. 438. The genus Rhizoctonus was erected for ampelinus Mok., a species occurring on tlie vine. Tlirough the kindness of H, F. Wilson the writer has been able to examine a slide containing alate forms. These, however, are in a very poor condition and it is impossible to determine whether or not cornicles are present. The antennae too are much distorted. This slide seems to indicate, however, that ampelinus is somewhat related to passerinii Sig., an hypothesis which is strengthened by the fact that both species hold the wings flat in repose. The genus, therefore, is placed here with some hesitation. Characters.— Antennse of six segments, without secondary sensoria and rather thick. Fore wings with the media simple. Hind wings with both media and cubitus present. Wings held flat in repose. Cauda and anal plate rounded. Tj^e fmonotn^ical), Rhizoctonus ampelinus Mok. Tribe MELAPHINI, The tribe Melaphini is suggestive both of the Pempliigini and of the Eriosomatini. It is, however, quite distinct from both. The habits more nearly resemble those of the Pem])higini. The tribe is placed here but a study of the sexes may show that it really belongs to the Hormaphidinae. The typical host group is Rhus, and the galls are typically fall galls. Characters. — Gall-inhabiting forms. Cornicles absent; antennse of the alate form of five or six segments armed with somewhat oval or linear transverse sensoria. Sexual forms not known. The genera may be separated as follows: Key to the Genera op the Melaphini. 1 . Both media and cubitus present in the hind wings 2. Only the media present in the hind -wings Aplonem-a. 2. Stigma of four wings pointed on distal portion and extending some distance Melaphis. Stigma stopiiing abruptly on distal extremity 3. 3. Antennse of five segments Nurudea. Antennae cf six segments Pemphigella. Genus APLONEURA. Plate XII, A-E. 1863. Aploncwrn Passerini, Aphididae Italicac, p. 78. 1869. Tctrcncma Derbes, Ann. des Se. Nat. Zool. (5), v. 11, p. 106. 1848. Baizongh Rondani, Nuovi .\nnali delle Scienze Natural!, v. 9, p. 35. The genus is distinguished quite easily from related one's by the venation of the hind wings, the relation of the cubitus and anal of the fore wings, and by the structure of the antenna\ The positive determi- nation of the insect of Fabricius maj^ cause A]:)loneura to fall for Rondani's name. 74 BULLETIN 826;, V. S. DEPARTMENT OF AGRICULTURE. Characters. — Cornicles absent. Stem-mother with five-segmented antennae; alate form with the media simple, the cubitus and anal joined near their bases. Hind wings with only the media present, antennae of six segments, armed with large subcircular or elongate sensoria. Forms living in true galls. Type (monotypical), Tctraneura lentici Pass. Genus MELAPHIS Walsh. Plate XII, F-K. 1866. Melaphis Walsh, Proc. Ent. Soc. Phila., v. 6, p. 281. 1883. Schlechiendalia Lichtenstein, Stett. Ent. Zeit., v. 44, p. 240. 1905. Abamalekia Del Guereio, Pvedia, v. 3, p. 364. Walsh erected his genus for rliois Fitch, a species formmg galls on sumach and in his discussion mentioned the Chinese gall, wonder- ing if it could be congeneric. The writer has recently published an account of the Chinese gall ^ and therein placed Schlechtendalia Licht. as a synonym of Melaphis. Del Guercio's genus was erected with his lazaretoi as type and placed as a thelaxine. Although the writer has never obtained specimens of this species he is of the opinion that it can not possibly be one of these insects and that it evidently belongs m the Melaphmi w^here the shape of the stigma would imme- diately place it as a Melaphis. Characters. — Cornicles absent. Apterous form with five-segmented antennse. Alate form with either five or six segmented antennas which are armed with some- what linear or oval sensoria. Fore wings with the media simple, although sometimes slightly forked, distal extremity of the stigma rather long drawn out; hind wings with both media and cubitus present; abdomen with distinct wax plates. Forms living in galls from which the alate forms escape in late summer or fall. Type (monotypical), Byrsocrypta rhois Fitch. Genus NURUDEA Matsumura. Plate XII, L-Q. 1917. NuTudea Matsumura, Synopsis of the Pemphigidae of Japan, p. 65. 1917. Nurudeopsis Matsumura, Synopsis of the Pemphigidae of Japan, p. 67. 1917. Fushia Matsumura, Synopsis of the Pemphigidae of Japan, p. 70. Matsumura erected his genus Nurudea for his Nurudea ihofushi, a species somewdiat similar to a species of Melaphis. This species differs quite markedly, however, in the form of the stigma. At the same time he erected the genus Nurudeopsis w^ith N. shiraii as tj^pe. This species differs little from ihofuslii exceptmg in the proportions of the segments, and hi the fact that the cubitus and first anal are some- what closer togetner at the base. The writer is of the opmion that these characters are not sufficient on which to form another genus. In the same w^ork also he erected the genus Fushia with FusJiia rosea Mats, as type. This species differs somew^hat from the type of Nuru- dea but the waiter believes that there are not differences sufficient to cause this to be considered as a separate genus. The antennae 1 Baker, A. C. On the Chinese gall (Aphididae— Ilom.). In Ent. News, v. 28, p. 385-393, 1917. GENERIC CLASSIFICATION OF APHIDIDAE. 75 are more slender and of somewhat different proportions and the cubitus and anal of the fore wmgs are united somewhat at base. In other genera, however, this difference between species is to be found, and it seems wisest not to adopt it in this group as of generic im- portance. Characters.— Cornicles absent. Antenuse of five segments armed with linear or somewhat oval sensoiia. Fore wings with the media simple, the stigma normal; hind wings with both media and cu1)itus present. Sexes unknown. Forms making galls upon the leaves of plants. Type (fixed by Matsumiira, 1917), Nurudea ibofushi Mats. Genus PEMPHIGELLA TuUgren. 1909. Pemphigclla Tullgren, Arkiv for Zool., v. 5, p. 171. 191S. Dasia Van der Goot, Mem. Ind. Mus., v. 6, p. 152. • Characters. — Cornicles absent. Antennae of six segments armed with oval sensoria. Hind wings with both media and cubitus present. Sexes unknown. Species form- ing galls on plants. Type (monotypical), Tctraneura cornicularia Pass. Tribe PROCIPHILINI. The tribe Procipliilini contains forms wliich specialized in some direc- tions more than did the Eriosomatini. In other ways, however, they appear to be more primitive than certain genera of that tribe. The cornicles have here disappeared altogether and large wax areas have replaced them. In their habits of gall formation, however, the Eriosomatini are more advanced than are members of the present tribe. Characters. — Forms living in crumpled or twisted leaves or in a somewhat complete gall caused by the rolling up of the leaves of the host. Wax plates present; cornicles absent; antennteof stem mother of five segments; those of the alate form six segmented and armed with narrow, transverse, or somewhat broadly oval sensoria. Key to the Genera of the Prociphilini. 1. Media once forked, stem mother usually without wax plates Asiphum. Media simple, stem mother wdth several rows of wax j^lates 2. 2. Wax plates large, those on the thorax well developed, stem mother and offspring living together 3. Wax plates not well developed, stem mother usually in a gall by herself . .Thecabius. 3. Sensoria narrow, linear, ciliate Prociphilus. Sensoria somewhat oval, nonciliate Neoprociphilus. Genus ASIPHUM Koch. Plate XU, R-X. 1857. Asiphum Koch, Die Pflanzenlause Aphiden, p. 246. 1859. Type fixation, Gerstaecker, Bericht for 18,57, p. 249. 1905. Type fixation, Kirkaldy, Can. Ent., v. 37, p. 418. The genus Asiphum was erected by Koch with two species, popwZi Fab. and ligustrinellum Koch. He listed De Geer's work in the literature under populi Fab. The populi of Fabricius proves to be the tremulae of De Geer. Only the species ligustrinellum was placed 76 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. in the genus by Lichtenstein (1885) and this species has been indicated as type of the genus by Kirkaldy (1905). Tremulae De Geer is a well- known species but the writer has been unable to obtain ligustrineUum and, in fact, has been unable to learn anything definite in regard to the species. The following conception of the genus, therefore, is based upon tremulae De Geer in view of the fact that ligustrineUum appears to be unknown, and since Apliis fajmli Fab. was indicated by Gerstaecker in 1859. Characters. — Cornicles absent; wax plates present in the alate forms; stem motlier ■with five-segmented antennae. AHate form with six-segmented antennae which are armed with rather narrow transverse sensoria. Fore wings with the media once forked^ hind WT.ngs with both media and cubitus present. Forms li\ing in the somewhat crumpled leaves of their host. Type (fixed by Gerstaecker, 1859), Aphis populi Fab. Genus NEOPROCIPHILUS Patch. Plate XIII, A-F. 1912. Neoprociphilus Patch, Bui. Me. Agr. Expt. Sta., no. 202, p. 174. The genus Neoprociphilus Patch is very close indeed to Prociphilus, the characters which separate it being the somewhat more oval or rounded sensoria and the fact that the sensoria are not ciliate. How- ever, in some of the species of Proci])liilus, particularly in the fall forms, somewhat oval sensoria are met with. It is retained doubt- fully. Characters. — Stem mother with five-segmented antennae. Cornicles absent, large wax plates similar to those of Prociphilus present. Alate form ■with six-segmented antennae which are armed with oval or subcircular nonciliate sensoria. Fore wings ■ndth the media simple, hind ■nings with both media and cubitus present. Sexes small, apterous, and beakless. Oviparous female developing only one egg. Forms living free upon their host, the stem mother and following generations in company. Ty]5e (monotypical), Pemphigus atlemudus O. S.. Genus PROCIPHILUS Koch. Plate XIII, G-N. 1857. Prociphilus Koch, Die Pflanzenlause, p. 279. 1857. Stagonia Koch, Die Pflanzenlause, p. 284. 1875. HoUncria Lichtenstein, Bui. See. Ent. Fr. (5) v. 5, p. lxxvi. 1917. NisMyana Matsumura, Synopsis of the Pcmphigidac of Japan, p. 90. In 1857 Koch erected his genus Prociphilus with three species: humeliae Schrank, erraticus Koch, and gnaplialii Kalt. Later in the work (p. 284) he used xylostei De Geer as the type of the genus Stagonia. This species, xylostei, is in all respects similar to humeliae and therefore Stagonia becomes a synonym of Prociphilus. In 1875 Lichtenstein erected the genus Holzneria with voscJiingeri Holzner as type. PoscTiingeri has been considered by many authoi^s as the alternate form of humeliae. In such case Holzneria must necessarily GENERIC CLASSIFICATION OF APHIDIDAE. 77 be a synonym of Prociphilus. Should poschingeri, however, be proven to be a distinct species it is so simihir in all regards that Holzneria must remain a synonym. In 1917 Matsumura erected the genus Nishiyana with N. aomori- ensis Mats, as type, placing it close to Prociphilus. From this genus he separated it because of the absence of wax plates in the thorax and the somewhat shorter antennae. It must be borne in mind, however, that the specimens he had were fall migrants. Fall mi- grants of several species of Prociphilus show very reduced wax plates, and in some these are absent altogether, although distinct in the spring migrants. It is believed that this genus is in reality Proci- philus. Characters. — Cornicles aosent, 'wax plates present, very large and well develoised. Stem mother with fi ve-segmented antennae. Alate form with six-segmented antennae armed with narrow transverse sensoria; secondary sensoria fringed. Fore wings with the media simple, hind wings with both media and cubitus present. Sexes small, apterous, and beakless. Oviparous females developing only one egg. Forms li\'ing in pseudogalls, the stem mother and her offspring together. TjT^e (Sxed by Gerstaecker, 1859), Aphis biuncUae Schr. Genus THECABIUS Koch. riate XIII, O-U. 1857. Thecabius Koch, Die Pflanzcnliiuse, p. 294. 1886. Bucktonia Lichtenstein, Monogr. d. puccrons, p. 16. The genus Thecabius was erected by Koch in 1857 for his species populneus. This species proves to be a synonym of Pemphigus affinis Kalt. -In 1886 Lichtenstein erected the genus Bucktonia with affinis Kalt. as type. Bucktonia, therefore, becomes a synonym. Characters. — Cornicles absent; wax plates present but not prominently developed as in Prociphilus. Stem mother with five-segmented antennae and rather narrow sensoria, secondary sensoria not fringed ; fore wings mth media simple, hind winga with both media and cubitus present. Sexual forms small, apterous, and beakless, oviparous female developing only one egg. Forms living in galls, the stem mother usually living in a gall by herself. Type (monotypical), Thecabius populneus Koch {=Pemphigus affinis Kalt.) This genus is closely related to Prociphilus and it is with some hesitation that the writer places it as distinct. Certain species, such as patcTiii Gillette, which are undoubtedly congeneric with affinis, do not show the typical life habit of the stem mother living in a gall alone. However, the character of the sensoria and the undeveloped nature of the wax glands ma}- serve to distinguish the genus. Tribe FORDINI. Members of the Fordini are specialized subterranean forms mostly living in the nests of ants. The aphids excrete honeydew, in return for which they are tended carefully by these insects. The apterous 78 BULLETIN 826, U. S. DEPAETMEISTT OF AGRICULTURE. forms generally are of a yellowish or brownish-yellow color, some- times a milk white. Wax-secreting plates are present but they are not developed to the same extent as are those of the Prociphilini. The cornicles are lost entirely and the region where these usually occur is occupied by wax plates. In some species the wax areas are reduced. Some species are armed with fine hairs, whereas others are almost entirely smooth. The eyes in the apterous forms are composed of three facets. The alate forms have rather short, thick antennae with somewhat oval sensoria. Three genera compose the tribe and these may be separated as follows : Key to the Genera of the Fordini. 1. Antenn£e of the alate form composed of five segments Forda. Antennae of the alate form composed of six segments 2. 2. Sensoria of the antennae of the alate form small and scattered over most of the segment, a central triangular wax plate on the thorax; apterous form with six-segmented antennae Paracletus. Sensoria of the antennae of the alate form larger and more evenly placed; the sensoria sometimes extending evenly across the segment. Apterous form with five-segmented antennee, sometimes with six segments present Geoica. Genus FORDA Heyden Plate J^III, V-AA. 1837. Forda Heyden, Mus. Sinkbg., v. 2, p. 291. 1841. Rhizoterus Hartig. Zeit. Ent., v. 3, p. 363, 1849. Smynthurodes Westwood, Gardener's Chron., p. 420. 1896. PentapMs Horvath, Wien. Ent. Zeit., v. 15, p. 2. 1909. PentapMs Del Guereio, Rivist. Patol. Vegetale, n. s., v. 3, p. 332. 1914. Eectimsus Theobald, The Entomologist, v. 47, p. 28. In 1837 Heyden erected his genus Forda, the type of which is formicaria Heyden. In 1896 Horvath erected his genus Pentaphis with marginata Koch as a type, while Del Guereio in 1909 used trivialis Pass, as the type of a genus of the same name. Specimens of marginata Koch from Horvath prove that this species in every respect is similar to the type of the genus. Pentaphis, therefore, will become a synonym of Forda. Likewise specimens of trivialis show that this species belongs in the same genus. In 1841 Hartig erected the genus Rhizoterus, the type of which is vacca. According to Lichtenstein this species is a synonym of formicaria Heyden, and Rhizoterus also, then, becomes a synonym. In 1914 Theobald erected his genus Rectinasus with his huxtoni as type. He based his genus on the proportions of the antennal segments, their length, and the length of the beak. The write; is opposed to basing genera on the proportions of the antennal segments, for in species in which these are of different proportions a very close relationship is evident. This is also true of the beak. Many Ameri- can species taken in ants' nests and as yet undescribed have beaks ranging from small to longer than the body, but they are all evidently GENEKIC CLASSIFICATIOlSr OF APHIDIDAE. 79 closely related. The other characters mentioned by Theobald are seen to be present in the type species. The tubercles he figures and describes are the same and the spines on the first and second antennal segments are evidently the thickened, pointed, chitinized articula- tions of the segments common in insects of this type. We believe, therefore, that Rectinasus should be carried as a synonym of Forda. Westwood's genus was erected on his hetae which appears to belong here as recently indicated by the writer. Characters. — Cornicles wanting; apterous forms with five-segmented antennae and eyes of three facets. Al'ate form with five-segmented antennae and medium-sized oval, or more or less irregularly shaped sensoria. Fore mngB with media simple; hind wings with both media and cubitus present, arising slightly apart. Subterranean forms living usually in the nests of ants and tended by them. Type (monotypical), Forda formicaria Heyden. Genus GEOICA Hart. Plate XIV, A-K. 1894. Geoica Hart, 18th Rept. State Ent. 111., p. 101. 1860. Tycliea Passerini, Gli Afidi, p. 30. 1906. Tycheoides Schouteden, Mem. Soc. Ent. Belg., v. 12, p. 194. 1906. KaltenbachkHa Schouteden, Mem. Soc. Ent. Belg., v. 12, p. 194. 1909. TrifidapMs Del Guercio, Rivista Patol. Vegetale, n. s., v. 3, p. 332. 1912. Tullgrenia V. d. Goot, Tijdschr. voor Ent., v. 15, p. 96. 1913. Trinacriella Del GuQrcio, Redia, v. 9, p. 1G9. 1916. Serrataphis V. d. Goot, Zur Kenntniss der Blattlause Java's, p. 263. In 1860 Passerini used the generic name Tychea of Koch and placed as the typical species i)h,aseoli Pass. In 1863 he used the name again, listing several species but not the species included by Koch. Therefore his interpretation of the genus can not be correct. In 1894 Hart erected his genus Geoica with squamosa as type. In 1906 Schouteden noticed Passerini 's mistake and suggested the name Tycheoides but made eragrostidis Pass, the type. In the same year he erected Kaltenbachiella with menffiae Schout. as type. In the year 1909 Del Guercio erected Trifidaphis with radicicola Essig as type. In 1912 Van der Goot noted Passerini's mistake and proposed the name Tullgrenia for the Tychea of Passerini. In 191 G Van der Goot erected the genus Serrataphis with lucifuga Zehntner as type. In studying cotypes and other specimens of squamosa, certain generic characters are evident. The species is subterranean. It has five-segmented antennse in the apterous form and six-segmented ones in the alate. It is true, however, that the apterous form sometimes has only four segments in the antennae and the alate five. Indeed, in some alate forms there is a five-segmented antenna on one side and a six-segmented one on the other. One of these five- segmented antennae was figured by Hart. One wmg vein only was figured in the hind wing by Hart, but a very close examination shows that both the media and cubitus, though faint, are present. These are very difficult to trace in balsam mounts. In giving his name 80 BULLETIN 826_, U. S. DEPARTMENT OF AGRICULTURE. Tycheoides Schouteden makes plain that he is naming the Tychea of Passerini and yet he sets a different type. He further states: ''Le genre Tycliea est vraisemblablement destine a desparaitre, ses especes appartenant en r^ahte a Tetraneura ou ByrsocryptaJ' He evidently is speaking here of the Tychea of Koch, since specimens of Tychea Pass, do not possess the cornicles of either of the two genera mientioned. In describing Kaltenbachiella Schouteden gives as a character the four-segmented antennte, etc., of the apterous form. The alate form he did not know, but from his description of the pupa it evi- dently would possess six-segmented antennae. There seems little doubt that this is another such case as squamosa where the apterous form has often four-segmented antennae, although five is the normal number, the alate form normally having six. Hart's description has led Schouteden astray and he places Geoica close to Forda, separating it therefrom by the venation of the hind wing. Consider- ing all of these facts there seems little doubt that Kaltenbachiella should be placed as a synonym of Geoica. Specimens of pJiaseoli show very similar characters in every respect. It is true that the antennae are somewhat longer and the liahs simple, but in every respect of importance the insects agree. The apterous form has five-segmented antennae and the alate form has six-segmented ones with the sensoria very similar in nature. The Cauda also is very similar. It is evident then that Tychea Pass, and Tullgrenia V. d. Goot become synonyms of Geoica Hart. Specimens of lucifuga Zehntner show a remarkable resemblance to squamosa with the exception, of course, of the squamae. The apterous form has five-segmented and the alate form six-segmented antennae. In general form and structure of the caudal extremity the insects are the same and, therefore, Serratapliis V. d. Goot will become a syno- nym of Geoica. There remains, then, to discuss the genus Trifidaphis Del Guercio. The cotypes of the type species show a close resem- blance to the general type of squojmosa. The apterous forms have five-segmented antennae and in the alate form, as in squamosa, some forms have five segments and some forms have six. The general • resemblance in other respects seems to prove that Trifidaphis is a synonym of Geoica. It should be pointed out that the sexes described by Hart are in all probability not sexual forms, but immature specimens. In 1913 Del Guercio erected the genus Trinacriella for his new species magnijica. Pie gave a brief description stating that the apterous forms had five-segmented antennas and the alate forms six. No specimens of this species are available to the writer, but there seems little doubt that Trinacriella will become a s3monym of Geoica. GENERIC CLASSIFICATION OF APHIDIDAE. 81 Characters. — ■Cornicles ^/autiug; apterous form usually with five-segmented antennae and eyes of three facets. With the intermediate forms more facets may occur. Alate form with usually six-segmented antennae and rather large oval sensoria with distinct rims. Fore wings with media simple. Hind wings with both media and cubitus present, though these may be faint and almost obscured in balsam. Cauda large and somewhat rectangular or roimded. Subterranean forms living on the roots of plants. Sexes small, apterous, and beakless. Type (monotypical), Geoica squamosa Hart, Genus PARACLETUS Heyden. Plate XrV, L-S. 1S37. Paradctus Heyden, Mas. Stnkbg., v. 2, p. 295. The genus Paracletus is closely related to Forda Heyden from, which it may be distinguished by the number of antennal segments in both alate and apterous forms. As with other genera of this tribe the eyes of the apterous form consist of three facets. Intermediate forms often occur, however, in which the intermediate nature is indicated only by the eyes which have the beginnings of compound eyes, such as are found in the alate form. There never are, however, complete compound eyes. The genus was erected by Heyden in 1837. The species live in close association with ants. CAf/racters.— -Cornicles absent. Apterous form with six-segmented antennae and eyes of three facets; alate form with six-segmented antennas wliich possess many rather small, oval sensoria. Fore wings with media simple, hind wings with both media and cubitus present, arising some distance apart. Thorax with a central wax plate. Forms living in the nest of ants and cared for by them. Sexes small, apterous, and beakless. In some cases only one claw is met with on the foot, while in other cases the normal number of two is present. This appears to be no definite character, as sometimes a claw is dropped from one foot and sometimes from another. Type (monotypical), Paracletus cimiciformis Heyden. Subfamily IV, HORMAPHIDINAE. The genera placed in this subfamily have usually been placed with the Eriosomatinae, or Pemphiginae, as it has been sometimes called. Mordwilko, however, placed these forms as his tliird tribe under the subfamily Aphidinae next to hJs tribe Callipterea. Something can be said in favor of both of these placings. In the first instance, the species in general form, antennal structure, and habit of gall formation are no doubt suggestive of the Eriosomatinae. On the other hand, their structure in regard to Cauda and anal plate is very like the Callipterina and the sexual forms appear to have a development of their own, although they are nearer in many ways to the Aphidinae than to the Eriosomatinae. It is the author's belief that these forms should constitute a sepa- rate subfamily. It has developed the habit of gall formation and the sensory characters which usually accom23any it, while at the same time it has retained in the sexual female the normal develop- 141613°— 20— Bull. 826 6 82 BULLETIN" 826, V. S. DEPARTMENT OF AGRICULTURE. merit of the ovaries found in the more primitive groups, and has retained in both sexual forms the beak and the ability to feed. This at once suggests a different line of development from that taken by the Eriosomatinae, although in some of its habits the Hormaphidinae agrees with that subfamily. In other lines, however, marked differences are met with here and one of the most striking of these is development of aleyrodiform generations, which remain stationary upon the host. Such a development is never met with in the Eriosomatinae, although the sexual forms are much more specialized. Since many of the genera of the Hormaphidinae are gall formers, sensoria very similar to those met with in the Eriosomatinae are met with here also. Indeed the same annular sensoria found in the Eriosomatini are even more pronounced in the Hormaphidinae and the sensoria on the wing bases are prominent and often, numerous. The cornicles in the present subfamily are sometimes absent or, as is usually the case, reduced to mere rings. In some genera, how- ever, they may be elevated sUghtly on broad shallow cones, somewhat suggestive of those of Anoecia. No. prominent cornicles, however, occur. In the wing venation there is often a considerable reduction and this shows also the specialized nature of the insects. The venation is comparable to that met with in the Eriosomatinae. In the fore wings the media is either simple or once branched, the radial sector, cubitus, and anal are present, but the cubitus and anal are often fused near their bases. In the hind wings both the media and cubitus are sometimes present, but often only the media remains. Great specialization in wax-producing organs occurs. In many of the forms these agree with the ones found in the Pemphigini. In certain aleyrodiform generations and in some sexual forms agglomerate glands or rather groups of small glands are seen. These may be arranged in different ways and often are placed about the margin of the insect so that it possesses a distinct lateral fringe, ' very like that of an aleyi'odid. In fact some of these insects on this account are very often mistaken for aleyrodids. The sexual forms are often quite small and possess large wax- producing areas. Others may lack these. All, however, develop to normal adults. The habit of gall formation is very marked here. Indeed, some species form galls on two different species of plants, migrating between the two. Characters. — Aerial forms living iu galls or sometimes free upon the host. The [ mesothorax in many forms altered so that its divisions are more or less miobservable, the entire mesothorax often showing as only one plate. Scalelike or aleyrodiform GENERIC CLASSIFICATIOlSr OF APHIDIDAE. 83 generations often developed. Cornicles often reduced to mere ringlike openings or entirely absent. Sexual forms small and apterous but with fully developed beaks. Oviparous female laying several eggs. Key to the Tribes of the Hormaphidinae. 1. Aleyrodiform generations developed 2. Aleyrodiform generations not developed Oregmini. 2. Cornicles absent; insects usually gall formers Hormaphidini. Cornicles usually present; insects usually not gall formers Cerataphidixi. Tribe HORMAPHIDINI. Members of this tribe are distinguished easily from those of other tribes in that the cornicles are absent and aleyrodiform generations are developed. These remain more or less quiescent upon the foliage. Some different forms of the species are often gall pro- ducers. All secrete wax from special pores. Considerable variation is met with in the development of the aleyrodiform generations. Sensoria of the alate forms are usually narrow and annular. Only two genera are so far recorded. Key to the Genera of the Hormaphidixi. Antennse of the alate forms composed of three segments, hind wings with only the media present Hormaphis. Antennse of the alate forms composed of five segments, hind wings with both media and cubitus present Hamamelistes. Genus HAMAMELISTES Shimer. Plate XIV, T-X. 1867. Hamamelistes Shimer, Trans. Am. Ent. Soc, v. 1, p. 284. 1896. TetrapMs Horvath, Wien. Ent. Zeit, v. 15, p. 6. Shimer included two species in this genus, spinosus Shimer and comu Shimer. The latter species, as suspected by him, is a syno- nym of Jiamamelidis Fitch. This species has been made the type of Hormaphis. Characters. — Cornicles absent. Stem mother with four-segmented antennse. Aley- rodiform generations developed. Alate form with five-segmented antennse which are armed with numerous annular sensoria. Wings held fiat in repose; fore wings with the media simple; hind wings with both media and'cubitus usually present; cauda knobbed, anal plate bilobed; wax -secreting areas abundantly present in the apterous forms. Sexes small and apterous but with beaks developed, oviparous female laying several eggs. Forms living in galls upon the leaves or scale-like on the leaves or twigs. Type (one unquestioned species), Hamamelistes spinosus Shimer. 84 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Genus HORMAPHIS Osten-Sacken. Plate XIV, Y-FF. 1861. HormapTiis Ostcn-Sacken, Stettiner Ent. Zeit., p. 422. The genus Hormaphis was erected by Osten Sacken for a species he described as hamamelidis. This species it later proved was the same described by Fitch as Byrsocrypta hamamelidis. The genus later was made to include spinosus Shimer, but the distinction between this genus and the one described by Shimer has been pointed out by Pergande.^ Characters. — Cornicles absent; alejrrodiform generations developed; wax glands very numerous; stem mother with three-segmented antennae; alate forms with three- segmented antennae which are armed with distinct annular sensoria. Wings held flat in repose; fore ^vings with the media simple; hind wings ^vith the cubitus absent. Sexual forms small and apterous, possessing beaks, oviparous female laying several eggs. Type (monotypical), Hormaphis hamameUdis O. S. {Byrsocrypta hamamelidis Fitch)- Tribe OREGMINI. Characters. — Forms living in galls cir otherwise upon the leaves of plants, possess- ing cornicles and Wax secreting glands. Antennae of the winged forms usually armed with annular sensoria; cauda rounded or somewhat knobbed, anal plate somewhat bilobed. No aleyrodiform generations developed. The sexual forms appear to be unknown. Key to the Gexera of the Oreomixi. 1. Vertex with two horn-like projections 2. Vertex without such horn-like projections 3. 2. Antennae five-segmented Oregma. Antennae foiir-segmented Ceratoglyphina. 3. Antennae of apterous form with five segments 4. Antennae of apterous form with four segments Glyphinaphis. 4. Media of fore wings once branched Astegopteryx . Media of fore wings simi)le Mansakia. Genus ASTEGOPTERYX Karsch. Plate XV, Q-X. 1S90. Astegopteryx Karsch, Ber. deutsch Botan. Ges., v. S, p. 52. 1906. Nipponapliis Pergande, Ent. News, v. 17, p. 205. 1916. Schizoncuraphis Van der Goot, Zur Kenntniss dcr Blattlausc Java's, p. 245. The genus Astegopteryx was erected with styracophila Karsch as type. Another species, nehoaslii, was described by Sasaki before the International Congress at Brussels in 1911 and a tliird species, styraci, was described by Matsumura in 1917. It is possible, there- fore, to gain a fair conception of the characters. In 1906 Pergande erected his genus Nipponaphis with distycMi Perg. as type. This species was stated to be from DistycJoium racemosum in Japan, on • Pergande, T. The life history of two species of plant lice inhabiting both the witch-hazel and birch. U. S. Dept. Agr. Bur. Ent. Tech. Ser. no. 9. 1901. GENERIC CLASSIFICATION OF APHIDIDAE. 8"5 which it forms galls. In 1916 Van der Goot erected his Schizoneura- phis with gallorum V. d. Goot as type. This species was said to form galls on Distylium steUare. The genus Astegopteryx can be separated as far as the recognized forms are concerned by the proportion of the antennal segments and some variation in the shape of the stigma. These chfferences are not, however, of large importance and Nipponaphis should be a synonym of Astegopteryx. In the same way, the type of Van der Goot's genus is not sufficiently different to warrant the erection of a new genus and Schizoneuraphis also should be considered asynonym. Characters. — Cornicles broad rings; apterous form with five-segmented antennae; alate form with five-segmented antennae which are armed with annular sensoria. fore wings with the media once l^ranched; hind wings with both media and cubitus present. Stigmal vein arising rather far back on the stigma. Cauda rounded, anal plate somewhat bilobed; forms living in galls. Sexual forms unknown. Type (mono typical), Astegopteryx styracophila Karsch. Genus CERATOGLYPHINA Van der Goot. I'late XV, M-P. 1916. Ceratoglyphina Van der Goot, Zur Kenntniss der Blattlause Java's, p. 237. Characters. — Cornicles present as mere pores. Vertex with two hornlike projec- tions. Antennaa of four segments; cauda and anal plate both rounded. Winged orms unknown . Type (fixed by V. d. Goot, 1916), Ceratoglyphina bambusaeY. d. Goot. Genus GLYPHINAPHIS Van der Goot. Plate XV, H-K. 1916. Ol'jphinaphis Van der Goot, Ziir Kenntniss der Blattlause Java's, p. 232. Characters. — Cornicles present as mere pores; antennte of four segments which are armed with linear sensoria. Fore wings with the media once branched; hind wings with both media and cubitus present; cauda knobbed; anal plate rounded; body covered with stout hairs. Type (fixed by V. d. Goot, 1916), Glyphinaphis bambusae V. d. Goot. Genus MANSAKIA Matsumura. I'late XV, L. 1917. Manmkia Matsumura, Synopsis of the Pemphigidae of Japan, p. 59. The author of the geaus Mansakia stated that it is closely allied to Hormaphis, but it would appear to the %vi'iter to be related .to the genera in the Oregmini as understood in the present classification. The presence of the cornicles would indicate that the genus is not related as closely to Hormaphis as to Astegopteryx, but its host and the nature of the gall would place it close to Hamamelistes. Since all of the forms are not known it is impossible to state positively its position. 86 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Characters. — Corniclea present as mere rings. Antennae of five segments armed with annular sensoria. Fore wings with the media simple, hind wings with two oblique veins. Cauda rounded, anal plate somewhat bilobed. Sexes not known. Forms causing galls upon plants, the galls usually armed with numerous long pro- jections. Type (fixed by Matsumura, 1917), Mansahia miyabei Mats. Genus OREGMA Buckton. Plate XV, A-G. 1893. Orcgma Buckton, Ind. Mias. Notes, v. 3, p. 87. 1897. Ceralovacuna Zehntner, Mededl. Proefs. Java, n. s., no. 37, p. 29. In 1893 Buckton established his genus Oregma with hambusae Buckton as type, while in 1897 Zehntner established his Ceratovacuna with lanigera Zehntner as type. Specimens of both of these species sent by Zehntner and specimens of hambusae from Green taken in Ceylon show that these genera must be considered the same. Characters. — Both alate and apterous forms with two hornlike projections on the vertex, wax gland areas present; antennae five-segmented, those of the alate form with narrow annular sensoria; cauda rounded or somewhat knobbed ; fore wings with media twice forked, hind winga with both media and cubitus present; cornicles l)road, slightly elevated rings. Type (monolypical), Oregma hambusae Buckt. Tribe CERATAPHIDINI. This tribe is closely related to the Hormaphidini but differs in that very distinct cornicles are here present. The apterous forms are scalelike and quiescent and feed upon the surfaces of the leaves. The alate forms possess annular sensoria. Wax secretion is abundant. The genera may be separated as follows: Key to the Genera of the Cerataphidini. Antenna) of the aleyrodif orm generations of five segments Aleurodaphia. Antennae of the aleyrodiform generations of four segments Cerataphis. Antennae of aleyrodiform genera tions of three segments Thoracaphia. Genus ALEURODAPHIS Van der Goot. Plate XVI, A-E. 1916. Aleurodaphis Van dor Goot, Ziir Kenntniss der Blattliiuse Java's, p. 239. This genus was erected for one species oc(turring in Java. Characters. — Form flat, aleyrodiform, three distinct divisions evident; cornicles present as mere rings. Margin with wax secreting glands; dorsum also with many small glands. Cauda rather elongate and knobbed, anal plate bilobed. Antennse five-segmented; eyes of the apterous forms with three facets. Type (fixed by Van der Goot, 1916), Aleurodaphis blumeae V. d. Goot. GENERIC CLASSIFICATION OF APHIDIDAE. 87 Genus CERATAPHIS Lichtenstein. Plate XVI, F-M. 1862. Bnisduvalia Signoret, Ann. Ent. Soc. France (4), v. 8, p. 400. 1882. Cerataphis Lichtenstein, Bui. Ent. Soc. France (6), v. 2, p. xvi. Signoret erected his genus Boisduvalia in connection with his aleyrodid monograph. He placed Coccus lataniae Bois. as type. Later, however, he considered this a coccid genus. The name was used in the Diptera in 1830, and is therefore not available. The name Cerataphis used by Lichtenstein in 1882 appears as the next name applied to the genus. Characters. — Cornicles present as mere rings. Apterous form with four-segmented antennae and aleyrodiform, with two divisions to the body; wax glands prominent; vertex with two hornlike projections; alate form with five-segmented antennae, the segments armed with narrow annular sensoria. Fore wings with the media once branched, hind wings' with both media and cubitus present. Cauda knobbed ; anal plate bilobed. Type (monotypical), Coccus lataniae Bois. Genus THORACAPmS Van der Goot Plate XVI, N-W. 1916. Thoracaphis Van der Goot, Ziir Kenntniss der Blattlause Java's, p. 242. Only the apterous form of one species of this genus has been described. Other species, however, are available for study, tlu'ough the generosity of Professor Van der Goot. Characters. — Cornicles present and quite distinct, occasionally absent, however, in the apterous form. Apterous form with three-segmented antennae, flat and with a posterior lobe. Alate form with five-segmented antennae armed with annular sen- soria. Fore wings with the media once branched ; hind wings with both media and cubitus present. Cauda somewhat knobbed, anal plate bilobed. Type (monotypical), Thoracaphis arboris V. d. Goot. 88 BULLETIT^ 826^ U. S. DEPARTMENT OF AGRICULTURE. GENERA NOT PLACED. A number of genera have been described which the writer has been unable to place. These genera are discussed in the following notes. Genus RHIZOEIUS Burmeister. 1835. Rhizobius Burmeister, Ilandbuch der Ent:oinolo,:;ie, p. 7.^. 1819. Rhizophlhiridum Van der Hoevcn, Handb. Dierkimde v. 1, p. 508. 1830. Rhyzoicus Passerini, Gli Afidi, p. 30. 1863. Rizobius Passerini, Aphididae Italicae, p. 79. 1919. Rhizoicus Del Guercio, Redia, v. 12, p. 251. The genus Rhizobius has generally been considered as a good aphid genus and writers have referred to species in this genus as having but one claw to the tarsus. However, as indicated under Paracletus the writer believes this is a variable character and we have no definite knowledge in regard to pilosellae Burm. Buckton's species of course was not in the original genus and therefore can not be used as type nor was it in Passerini's conception of Rhyzoicus. After placing Rhyzoicus Pass, with jujuhae Buckton as type, Del Guercio erects the genus Neorhizobius, distmguished by having two claws, and in which he places graminis Thos., poae Del Guercio, stramineus Del Guercio, and ulmipJiilus Del Guercio. In 1860 Passerini set sonclii Pass, as the type of Rhizobius Burm., and in a footnote suggested the name Rhyzoicus as a new name for Rhizobius, since this name had previously been used m the Coleoptera. Such procedure, however, is not allowable since sonchi Pass, was not in the original genus. Of the two species in the original genus inloseTlae Burm. has been accepted as type. Del Guercio in 1917 used the generic name Rhizoicus Pass., spelling it with an ''i" instead of a "y", and jujuhae Buckton as the tj^pe. In the writer's opinion the genus Rhizobius must remain unknown until the type species inlosdlae becomes known and carefully studied. The name Rhizophthiridum was given to this genus to replace Rhizobius Burm. Genus NEORHIZOBIUS Del Guercio. 1917. Neorldzobiwi Del Guercio, Redia, v. 12, p. 251. As indicated under the discussion of Rhizobius, four species are placed in this genus by Del Guercio. Three species are described as new and only in the apterous forms. Two of these forms have five-segmented antenna? and one of them four-segmented ones, and when the alate forms are found they will in all probability be shown to represent species either of Forda or Geoica. The basing of genera upon the relative lengths of the antennal segments would create a very large number of genera and separate related forms. The genus Neorhizobius, therefore, the r GEISTEEIC CLASSIFICATION OF APHIDIDAE. 89 writer considers composed of apterous forms, which really belono- in older and well-recognized genera, but which can not be definitely placed until the alate forms have been secured and studied. Genus SCHOUTEDENIA Riibsaamen. 1905. Schoutcdrnia Riibsaamen, Marcellia, v. i, p. 10. This genus, which was described for ralumensis Riib., is here listed as unknown. The gall formed by the species is described but the writer, never having been able to obtain either gall or insect, would be able only to guess at its position from the description. Genus CLAVIGERUS Szepligeti. Clavigerus Szepligeti, 1\ ovarAszati Lapok, v. 1, p. 4. This genus was described in the only volume issued of the Journal cited. The writer has been unable to secure a copy or any details of the description given. ADDENDA. The following generic names have been employed by Mordwilko (Fauna Russ. vol. 1, Aphidodea) without, apparently, the mention of any species in connection therewith. They have not been con- sidered in the foregoing paper: Anameson, Aorison, Chaetosiphon, Corylobium, Elatobium, Euaulax, Eurythaphis, Halmodaphis, Impatientinum, Jaxart aphis, Nasonovia, Orobion, Paczoskia, Sitobium, Staticobium, Tlja, Turanaphis, Uroleucon (subgenus), Uromelan. The subgenera Dact^aiotus Raf., Cladoxus Raf., and Adactynus Raf., have not been considered. INDEX TO GENERA. rage. Abamalekia Del Guercio 74 AVjiira Malsumura -13 Acanthaphis Matsumura 54 Acanthocallis Matsumura 29 Acaudus Van der Goot 41 Acyrthosiphon Mordwilko 56 Akkaia Takahashi 53 Aleurodaphis Van der Goot 86 Amphorophora Buckton 54 Amycla Koch 71 Anoecia Koch 13 Anomalaphis Baker 52 Anuraphis Del Guercio 42 Aphioides Passerml 36 Aphis lamiaeus 43 Aphoides Rondani 71 Aploneura Passerini 73 Arakawana Matsumura 34 Arctaphis "Walker 33 Arimakia Matsumura 43 Aristaphis Kirkaldy 36 Asiphum. Koch 75 Asiphonaphis Wilson & Davis 44 Aspidaphis Gillette 44 Astegopteryx Karsch 84 Atarsos Gillette 44 Atheroides Haliday 33 Aulacorthum Mordwilko 57 Baizongia Rondani 73 Boisduvalia Signoret 87 Brachycaudus Van der Goot 42 Brachycolus Buckton 45 Brachysiphum Van der (root 45 Brachyunguis Das 49 Bradyaphis Mordwilko 32 Brevicoryne Van der Goot 45 Brysocrj-pta Westwood 71 Bucktonia Lichtenstein 77 Byrsocrypta Haliday 71 Byrsocrypta Tullgren 67 Calaphis AValsh 26 Callaphis Walker 27 Callipterinella Van der Goot 26 CalUpteroides Mordwilko 28 Callipteroides Van der Goot 28 Callipterus Koch 27 Callipterus Van der Goot 29 Capitophorus Van der Goot 55 Carolinaia Wilson 45 Cavariella Del Guercio 46 Cerataphis lyichtenstein 87 Ceratoglyphina Van der G oot So Ceratovacuna Zelintner 86 Cerosipha Del Guercio 4fi Cervaphis Van der Goot 53 Ceylonia Buckton 51 Ohaitophorinella Van der Goot 34 Page. Chaitophorus Koch 33 Chelymorpha Clark 34 Chromaphis Walker 27 Cinaria Curtis 15 Cladobius Koch 36 Clavigerus Szi^pligeti 89 Colopha Monell 66 Coloradoa Wilson ^9 Cornaphis Gillette 09 Corynosiphon Mordwilko 46 Cryptosiphum Buckton 46 Dasia Vander Goot 75 Davisia Del Guercio 17 Dentatus Van der Goot 42 Dielcj'smura Mordwilko 56 Dilachuus Baker I6 Drepanaphis Del Guercio 31 Drepaniella Del Guercio 57 Drepanosiphum Koch 32 Dryaphis Del Guercio IS Dryoljius Koch 18 Dryopeia Kirkaldy 69 Eichochaitophorus Essig 33 Endeis Koch 69 Eriosoma Leach 66 Essigella Del Guercio 14 Eucallipterus Schouteden 28 Euceraphis Walker 28 Eudcis Ashmead 69 Eulachnus Del Guercio IS Eunectarosiphon Del Guercio 54 Eutrichosiphum Essig & Kuwana 38 Forda Heyden 78 Francoa Del Guercio 55 Fullawaya Essig 37 Fullawayella Del Guercio 59 Fushia Matsumura 74 Geoica Hart 79 Georgia Wilson 67 Glj-phina Koch 21 Glyphinaphls Van der Goot 85 Gobaishia Matsumura 67 Greenidea Schouteden 37 G reenidea Wilson — 38 Greenideoida Van der Goot 38 Hamadryaphis Kirkaldy 71 Hamamelistes Shimer 83 Hannabura Matsumura 41 Hayhurstia Del Guercio 47 Heteroneura Davis 47 Holzneria Lichtenstein 76 Hormaphis Osten Sacken 84 Hyadaphis Kirkaldy 47 Hyalopteroides Theobald 56 Hyalopterus Koch 47 Hysteroueura Davis 47 91 92 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. Idiopterus Davis 60 lUinoia Wilson 56 Kallistaphis ICirkaldy 28 Kaltenbaehiella Schouteden 79 Kesslcria Lichtenstein 71 Lachniella Del Guercio 15 Lachnus Burmeister 15 Liosomaphis Walker 48 Longicaiidiis Van dcr Goot 49 Longist igma Wilson 17 Longiunguis Van der Goot 43 Lowia Lichtenstein 72 Loxerates Raflnesquc 43 Maechiatiella Del Guercio 41 Macrosiphon Del Guercio 57 Macrosiphoniella Del Guercio 50 Maerosiphum Del Guercio 54 Macrosiphum Oestlund 54 Maerosiphum Passerini 57 Macrosiphum Van der Goot 56 Mansakia Matsumura 85 Mastopoda Oestlund 48 Megoura Buckton 57 Melanaphis Van der Goot 43 Melaphis Walsh 74 Mclanoxantherium Schouteden 36 Melanoxanthus Buckton 35 Metapliis Matsumura 4G Metopeurum Mordwilko 56 Micrella Essig 33 Micromyzus Van der Goot 59 Microparsus Patch GO Microsiphon Del Guercio 43 Microsiphum Cholodkovsky 49 Mimaphidus Rondani 06 Mindarus Koch 02 Monaphis Walker 32 Moncllia Oestlund 29 Morwilkoja Del Guercio 70 Myzaphis Van der Goot 43 Myzocallis Passerini 29 Myzodes Mordwilko 57 Myzoides Van dcr Goot 57 Myzopsis Matsumura 57 Myzoxylus Blot CO Myzus Passerini 57 Nectarophora Oestlund 57 Nectarosiphon Schouteden 54 Neocallipterus Van der Goot 28 Neomyzus Van dcr Goot 57 NeophyOaphis Takahashi 24 Neoprociphilus Patch 70 Neorhizobius Del Guercio SS Ncosymydobius Baker 32 Neothomasia Baker 35 Neotoxoptera Theobald 60 Neotrama Baker 20 Nippocallis Matsumura 27 Nippolachnus Matsumura 14 Nipponaphis Pergande 84 Nipposiphum Matsumura 46 Nishiyana Matsumura 76 Nurudea Matsumura 74 Nurudeopsis Matsumura 74 Oedisiphum Van der G oot 45 Oregma Buckton 86 Page. O vatus Van der Goot 57 Pachypappa Koch 71 Pachypappa Tullgren 71 Pachypappella Baker 71 Panaphis Kirkaldy 27 Paracletus Heyden 81 Patchia Baker 34 Pemphigella Tullgren 75 Pemphigus Hartig 71 Pentalonia Coquerel 61 Pentaphis Del Guercio 78 Pentaphis Horvath 78 Pergandeidia Schouteden 49 Periphyllus Van der Hoeven 34 Phillophorus Thornton 34 Phloeomyzus Horvath 72 Phorodon Passerini 58 Phyllaphis Koch 24 Phymatosiphum Davis 31 Prociphilus Koch 76 Protolachnus Theobald 15 Protrama Baker 19 Pterocallis Passerini 29 Pterochlorides Archangelsky 18 Pterochlorus Rondani 18 Pterocomma Buckton 36 - Ptychodes Buckton 27 Rectinasus Theobald 78 Rhizoberlesia Del Guercio 49 Rhizobius Burmeister 88 Rhizoctonus Mokrzecky 73 Rhizoicus Del Guercio 88 Rhizoicus Passerini 88 Rhizomaria Hartig 71 Rhizophthiridum Van der Hoeven 88 Rhizoterus Hartig 78 Rhopalosiphum Koch 49- Rhopalosiphum Passerini 57 Rhopalosiphum Van der Goot 54 Rhopalosiphon Scudder 49 Rhopalosiphontnus Baker 58 Rhynchocles Altum 18 Rizobius Passerini 88 Saltusaphis Theobald 31 Sanbornia Baker 50 Sappaphis Matsumura 42 Schizodryobius Van der Goot IS Schizolachnus Mordwilko 16 Schizoneura Hartig 66 Schizoneuraphis Van der Goot 84 Schlechtendalia Lichtenstein 74 Schoutedenia Riibsaamen 89 Semiaphis Van der Goot 42' Scrrataphis Van der Goot 79 Setaphis Van der Goot o9 Shivaphis Das , 24 Sipha Passerini 35 Siphocoryne Passerini 47,49 Siphonaphis Van der Goot 49 Siphonalrophia Swain 51 Siphonocallis Del Guercio 26 Siphonophora Koch 57 Smynthurodes West woo 1 78 Stagonia Koch l(x Stenaphis Del Guercio 43 Stephensonia Das 49 GENERIC CLASSIFTCATION OF APHIDIDAE. 93 I'age. Stomaphis Walker 18 Subcalliptenis Mord wilko 29 Symydobius Mordwilko 30 Takecallis Matsmniira 29 Tamalia Baker 24 Tetraneura Hartig 08 Tetraphis Horvath 83 Tet renema Derb63 T3 Thecabius Koch 77 Thelaxes Westwood 21 Therioaphis Walker 28 Thomasia WOson 35 Thoracaphis Van der Ooot 87 Thripsaphis Gillette 30 Todolachnus Matsumura 15 Toxoptera Koch 51 Trama Heyden 19 Tranaphis Walker 33 Travaresiella Del Guercio 21 Trichosiphum Pergande 37 Page. Trifldaphis Del Guercio 79 Trinacriella Del Guercio 79 Tuberculatus Mordwilko 29 Tuberculoides Van der Ooot 29 Tuberodryobius Das is Tuberolachnus Mordwilko ig Tullgrenia Van der Goot 79 Tychea Koch 71 Tychea Passerini 79 Tycheoides Schouteden 79 Unilachuus Wilson 17 Uraphis Del Guercio 43 Vacuna Heyden 2I Vesiculaphis Del Guercio 51 Watabura Matsumura 59 Wilsonia Baker 15 Yamataphis Matsumura 51 Yezabura Matsumura 42 Yezocallis Matsiunura 30 Yezosiphum Matsumura 49 PLATE I. A. — Anoecia querci, apterous form. B. — Anoecia querci, wings. C. — Anoecia querci, antenna of alate form. D. — Anoecia corni, fore wing. E. — Anoecia corni, antenna of alate form. F. — Anoecia corni, cornicle of alate form. G. — Nippolachnus pyri, fore wing. H. — Nippolachnus pyri, head of alate form. I. — Anoecia corni, head of alate form. J. — Nippolachnus pyri, cornicle of alate form. K. — Nippolachnus pyri, antenna of alate form. L. — Eulachnus agilis, body of alate form. M. — Eulachnus agilis, fore wing. N. — Eulachnus agilis, antenna of alate form. 0. — Eulachnus agilis, cornicle of alate form. P. — Eulachnus rileyi, head of alate form. Q. — Eulachnus rileyi, cornicle of alate form. R. — Eulachnus rileyi, segment III, antenna of alate form. S. — Essigella calif ornica, body of alate form. T. — Essigella calif ornica, fore wing. U. — Essigella californica, antenna of alate form. V. — Essigella californica, cornicle of alate form. W. — Essigella californica, head of alate form. X. — Essigella californica, cauda of alate form. Y. — Essigella californica, antenna of apterous form 94 Bui. 826. U. S. Dept. of Agriculture. Plate I. Bui. 826, U. S. Dept. of Agriculture. Plate 1 1 , PLATE II. A. — Dilachnus ponderosae, fore wing. B. — Dilachnus ponderosae, cornicle of apterous form. C. — Dilachnus ponderosae, head of alate form. D. — Schizolachnus tomentosus, fore wing. E. — Unilachnus parvus, fore Aving. F. — Unilachnus parvus, cornicle of alate form. G. — Unilachnus parvus, head of alate form. H. — Longistigma caryae, wings. I. — Longistigma caryae, apterous form. J. — Longistigma caryae, cornicle. K. — Longistigma caryae, cornicle. L. — Longistigma caryae, antenna of alate form. M. — Stomaphis quercus, alate form. N. — Stomaphis quercus, head of alate form. 0. — Stomaphis quercus, cornicle of alate form. P. — Stomaphis quercus, antenna of alate form. Q. — Stomaphis quercus, cauda and anal plate. P. — Stomaphis quercus, pits on rostrum. S. — Pterochlorus roboris, wing. T. — Pterochlorus roboris, head of alate form. U. — Pterochlorus roboris, cornicle of alate form. V. — Pterochlorus viminalis, wings. W. — Pterochlorus viminalis, abdomen of alate form. X. — Pterochlorus viminalis, head of alate form. 95 PLATE III. A. — Thelaxes dryopJdla, apterous form. B. — Thdaxes dryophila, wings. C. — Thelaxes dryopJdla, cornicle of alate form. D. — Thelaxes dryophila, cauda of alate form. 'E.— Thdaxes dryophila, cauda of apterous form. F.— Thelaxes dryophila, antenna of alate form. G. — GlypUna betulae, apterous form. H. — Glyphina betulae, wings. I. — Glyphina betulae, antenna of alate form. J. — Glyphina betulae, cornicle of alate form. K.— Glyphina betulae, cauda and anal plate of alate form. L,— Glyphina betulae, cauda and anal plate of apterous form. M. — Neotrama delguercioi, apterous form. N. — Trama troglodytes, apterous form. O. — Neotrama delguercioi, cornicle. P. — Prolrama radicis, apterous form. Q. — Protrama radicis, fore wing. R. — Protrama radicis, cornicle of alate form. S. — Protrama radicis, antenna of alate form. T". — Protrama radicis, tarsus of alate form. 96 Bui. 826, U. S. Dept. of Agriculture. Plate III. Bui. 826, U. S. Dept. of Agriculture. Plate IV. PLATE IV. A. — Amphorophora riibi, head of apterous form. B. — Amphorophora rubi, cornicle. 0. — Amphorophora rubi, cauda. D. — Rhopalosiphoninus laty siphon, head of alate form. E. — Rhopalosiphoninus laty siphon, cornicle of alate form. "F .—Rhopalosiphoninus laty siphon, cauda of alate form. G. — Myzocallis coryli, cornicle. H. — Myzocallis coryli, cauda and anal plate. I. — Callipterus juglandis, cornicle. J.—Callipterus juglandis, cauda and anal plate. K. — Therioaphis tiliae, cornicle. L. — Therioaphis tiliae, cauda and anal plate. M. — Monellia caryella, cauda and anal plate. N. — Monellia caryella, cornicle. O. — Chromaphis juglandicola, cornicle. P. — Chromaphis juglandicola, cauda and anal plate. Q. — Eucer aphis betulae, cauda and anal plate. R. — Eucer aphis betulae, cornicle. S. — Calaphis betulella, cauda and anal plate. T.—Calaphis betulella, cornicle. U. — Calaphis betulella, fore wing. V. — Saltusaphis scirpus, head of apterous form. W .—Saltusaphis scirpus, cauda and anal plate. X. — Thripsaphis balli, head of apterous form. Y. — Neothomasia populicola, cornicle. Z. — Neothomasia populicola, cauda and anal plate. AA. — Periphyllv^ negundinis, cauda and anal plate. BB . — Periphyllus negundinis, cornicle. CC. — Clmitophorus populi, cauda and anal plate. DD. — Symydobius oblongus, cornicle. EE. — Symydobius oblongus, cauda and anal plat' . FF. — Phyllaphis fagi, cauda and anal plate. GG . — Phyllaphis fagi, cornicle . HII. — Tamalea coweni, cornicle. II. — Tamalea coweni, cauda and anal plate. JJ. — Drepanaphis acerifolii, oviparous abdomen. IvK. — Drepanaphis acerifolii, cauda and anal plate of alate form. LL. — Drepanaphis acerifolii, cornicle. MM. — Drepanosiphum platanoides, cornicle. NN. — Melanoxantherium populifoliae, cornicle. 00. — Melanoxantherium populifoliae, cauda and anal plate. PP. — Pterocomma populeus, cornicle. 141613°— 20— Bull. 826 7 97 PLATE Y. A. — Eutrichosiphuin pasaniae, antenna of alate form. B. — Eutrichosiphum pasaniae, wings. C. — Eutrichosiphum pasaniae, cornicle of apterous form. D. — Eutrochosiphum pasaniae, cornicle of alate form. E. — Eutrichosiphum pasaniae, antenna of apterous form, F. — Greenidea anonae, apterous form. G. — Greenidea artocarpi, wings. H. — Greenidea artocarpi, cauda of apterous form. I. — Greenidea artocarpi, cornicle of apterous form. J. — Greenidea anonae, abdomen of alate form. K. — Greenidea anonae, head of alate form. L. — Greenideoida elongata, wings. M. — Greenideoida sp., head of alate form. N. — Greenideoida sp., cornicle of alate form. 0. — Greenideoida sp., third antennal segment of alate formT^ P. — Greenideoida hannae, cornicle of apterous form. Q. — Setaphis luteus, caudal portion of apterous form. R. — Setaphis luteus, wings. S. — Setaphis luteus, head of alate form. T. — Setaphis luteus, cornicle of alate form. U. — Setaphis luteus, spine of alate form. V. — Setaphis luteus, cornicle of apterous form. W. — Setaphis luteus, antenna of alate form. X. — Setaphis luteus, antenna of apterous form. 98 3ul. 826, U. S. Dept. of Agriculture Plate V. Bui. 826, U. S. Dept. of Agriculture. Plate VI. o PLATE VI. A. — Acavudus lychnidis, cornicle of alate form. B. — Acaudus lychnidis, caiida of alate form. C. — Anuraphis amygdali, cornicle of alate form. D. — Anuraphis amygdali, caiida of alate form. E. — Anuraphis amygdali, head of alate form. F. — Anuraphis carotae, cornicle and caiida of alate form. G. — Aphis sambuci, head of alate form. H. — Aphis sambuci, cornicle of alate form. I. — Aphis sambuci, cauda of alate form. J. — Brevicoryne brassicae, cauda of alate form. K. — Brevicoryne brassicae, cornicle of alate form. L. — Aspidaphis polygonii, cornicle of alate form. M. — Aspidaphis polygonii, cornicle more enlarged. N. — Aspidaphis polygonii, cauda of apterous form. 0. — Aspidaphis polygonii, head of apterous form. ^P. — Atarsos grindeliae, cornicle of alate form. Q. — Atarsos grindeliae, head of alate form. R. — Atarsos grindeliae, cauda of alate form. S.— Atarsos grindeliae, tibia of alate form. T. — Brachycolus stellariae, cornicle of alate form. U. — Brachycolus stellariae, cauda of alate form. V. — Carolinaia cyperi, cornicle of alate form. W. — Carolinaia cyperi, cauda of alate form. X. — Cavariella pastinacae, cornicle of alate form. Y. — Cavariella pastinacae, cauda of alate form. Z. — Cavariella pastinacae, tubercle of alate form. AA. — Hyadaphis xylostei, cauda of apterous form. BB .—Hyadaphis xylostei, cornicle of apterous form. CC. — Toxoptera aurantiae, head of alate form. DD. — Toxoptera aurantiae, cornicle of alate form. EE. — Toxoptera aurantiae, cauda of alate form. FF. — Rhopalosiphum rufomaculata, cornicle of alate form. GG. — Rhopalosiphum rufomaculata, head of alate form. HH. — Rhopalosiphum rufomaculata, cauda of alate form. II. — Rhopalosiphum nymphaeae, cornicle of alate form. JJ. — Rhopalosiphum nymphaeae, cauda of alate form. KK.—Pergandeidia ononidis, cauda and cornicle of alate form. LL. — Pergandeidia trirhodus, cauda of apterous form. MM. — Pergandeidia trirhodus, cornicle of apterous form. NN. — Liosomaphis berberidis, cornicle and cauda of apterous form. 00. — Liosomaphis berberidis, cauda of alate form. PP. — Cryptosiphum artemesiae, cornicle of apterous form. QQ. — Cryptosiphum artemesiae, cauda of apterous form. RR. — Hyalopterus arundinis, cornicle of alate form. SS. — Hyalopterus arundinis, cauda of alate form. TT". — Hyalopterus deformans, cauda of apterous form. UU. — Hyalopterus deformans, cornicle of apterous form. VV. — Hyalopterus deformans, cauda of alate form. WW. — Hyalopterus deformans, cornicle of alate form. 99 PLATl'. VII. A. — Siphonatrophia cupressi, apterous form. B. — Siphonatrophia cupressi, Avings. C. — Siphonatrophia cupressi, antenna of alate form. D. — Siphonatrophia cupressi, cauda of alate form. E. — Siphonatrophia cupressi, antenna of apteious form. F. — Sanhornia juniperi, apterous form. G. — Sanhornia juniperi, ■wings. H. — Sanhornia juniperi, antenna of alate form. I. — Sanhornia juniperi, head of apterous form from beneath. J. — Sanhornia juniperi, cauda of alate form. K. — Sanhornia juniperi, cornicle. L. — Sanhornia juniperi, tarsus. M. — Vesiculaphis caricis, apterous form. N.- — Vesiculaphis caricis, cornicle of alate form. O. — Vesiculaphis caricis, fore Aving. P. — Vesiculaphis caricis, antenna of alate form. Q. — Vesiculaphis caricis, cauda of alate form. R. — Acanthaphis rubi, apterous form. S. — Acanthaphis ruhi, head of apterous form. T. — Acanthaphis ruhi, cornicle of apterous form. U. — Acanthaphis ruhi, antenna of apterous form. 100 Bui. 826, U. S. Dept. of Agriculture. Plate VII. Bui. 826, U. S. Dept. of Agriculture. Plate VI 1 1. PLATE VIII. A. — Capitophorus shepardiae, cornicle of alate form. B. — Capitophorus shepardiae, head of alate form. C. — Capitophorus shepardiae, cauda of alate form. D.- — Anomalaphis comperi, apterous form. yj.—Anomalaphis comperi, antenna of apterous form. F. — Anomalaphis comperi, wings. G. — Cervaphis schoutedenine, extremity of abdomen. H. — Illinoia liriodendri, head of alate form. I. — Illinoia liriodendri, cornicle of alate form. J. — Illinoia liriodendri, cauda of alate form. K. — Myzus cerasi, head of apterous form. L.^ — Myzus cerasi, cauda of apterous form. M. — Myzus cerasi, cornicle of alate form. N. — Phorodon humuli, head of apterous form. O. — Phorodon humuli, cauda of apterous form. P. — Phorodon humuli, cornicle of apterous form. Q. — Phorodon humuli, head of alate form. R. — Macrosiphonella sanbomi, head of alate form. S.—Macrosiphonella sanbomi, cornicle of alate form. T. — Macrosiphonella sanbomi, cauda of alate form. IT. — Macrosiphum rosae, head of alate form. V. — Macrosiphum rosae, cauda of alate form. W.- — Macrosiphum rosae, cornicle of alate form. X. — Myzus persicae, head of apterous form. Y. — Myzus persicae, cornicle of alate form. Z. — Myzus pei'sicae, cauda of alate form. AA. — Microparsus variabilis, cauda of alate form. BB. — Microparsus variabilis, wings. CC. — Microparsus variabilis, cornicle of alate form. DD.— Microparsus variabilis, head of alate form. EE. — Idiopterus nephrolepidis, cornicle of alate form. FF. — Idiopterus nephrolepidis, fore wing. GG. — Idiopterus nephrolepidis, head of alate form. HH. — Idiopterus nephrolepidis, cauda of alate form. II. — Pentalonia nigronervosa, head of apterous form. JJ. — Pentalonia nigronervosa, wings. KK. — Pentalonia nigronervosa, cauda of alate form. LL. — Pentalonia nigronervosa, head of alate form. MM. — Pentalonia nigronervosa, cornicle of alate form. 101 Plate IX. A. — Mindarus abietinus, wings. B. — Mindarus abietinus, fore wing showing tracheae, C. — Mindarus abietinus, cauda of alate form. D. — Mindarus abietinus, antenna of alate form. E. — Mindarus abietinus, abdomen of pupa. F. — Mindarus abietinus, cornicle of alate form. G. — Colopha ulmicola, antenna of alate form. H. — Colopha ulmicola, antenna of stem mother. I. — Colopha ulmicola, wings. J. — Colopha ulmicola, apterous form. K. — Colopha ulmicola, antenna of apterous form. L. — Colopha ulmicola, antenna of apterous form. M. — Eriosoma lanigeruTu, apterous form. N. — Eriosoma lanigerum, wax plate. O. — Eriosoma lanigerum, wax reservoir. P. — Eriosoma lanigerum, fore wing. Q. — Eriosoma lanigerum, wing pad showing tra.— Aploneura lentici, wings. E. — Aploneura lentici, abdominal wax plate. F.—Melaphis rhois, antenna of apterous form, G. — Melaphis rhois, antenna of alate form. H.—IIelaphis rhois, wings. I. — Melaphis rhois, abdomen of alate form. J. — -Melaphis rhois, antenna of alate form. K. — Melaphis chinensis, fore wing. L. — Nurudea ibofushi, wings. M. — Nurudea ibofushi, antenna of alate form, N. — Nurudea shiraii, wings. O.—Nurudea shiraii, antenna of alate form. P. — Nurudea rosea, fore wing. Q. — Nurudea rosea, antenna of alate form. II . — Asiphum tremulae, stem mother. S. — Asiphum tremulae, antenna of stem mother. T. — Asiphum tremulae, antenna of alate form. U. — Asiphum tremulae, wings. Y .—Asiphum tremulae, abdomen of alate form. W. — Asiphum tremulae, wax plates. X. — Asiphum tremulae, antenna of apterous form. 105 PLATE XIII. A. — Neoprociphilus attenuatus, fore wing-. B. — Neoprociphilus attenuatus, antenna of alate form. C. — Neoprociphilus attenuatus, segment V of antenna of alate foroi* D. — Neoprociphilus attenuatus, antenna of apteroiie form. E. — Neoprociphilus attenuatus, antenna, of male.. F. — Neoprociphilus attenuatus, antenna of oviparous femaJte. G. — Prociphilus bumeliae, antenna of alate female. H.^Prociphilus bumeliae, antenna of stem; mother. I. — Prociphilus bumeliae, wings. J. — Prociphilus bumeliae, abdomen of alate form. K. — Prociphilus bumeliae, thoracic wax plates: L. — Prociphilus bumeliae, head wax plates. M. — Prociphilus xylostei, thoracic wax plates. N. — Prociphilus bumeliae (poscheringi) , thoracic wax pMiasi,. O. — Thecabius affinis, antenna of stem mother.. P. — Thecabius affinis, antenna of alate form. Q. — Thecabius affinis, wings. R. — Thecabius affinis, thorax of alate form. S. — Thecabius affinis, head wax plates of stem mother.. T. — Thecabius affinis, abdominal wax plate. U. — Thecabius affinis, abdominal waaE plia*e&of atem. mother. V. — Forda sp., apterous form. W. — Forda olivacea, antenna of alate form. X. — Forda trivialis, antenna of alate form. Y. — Forda marginata, antenna of alate form. Z. — Forda olivacea, fore wing. AA. — Forda formicaria, antenna of apterous form. 106 Bui. 826, U. S. Dept. of Agriculture. PLATE XMI. Bu!. 826, U. S. Dept. of Agriculture. Plate X I V. PLATE XIV. A. — Geoica squamosa, apterous form. B. — Geoica squamosa, antenna of alate form. C. — Geoica squamosa, eye of apterous form. D. — Geoica squaviosa, wings. E. — Geoica squamosa, caudal extremity of apterous form. F. — Geoica squamosa, squama. G. — Geoica squamosa, four-segmented antenna of apterous form. H. — Geoica phaseoli, antenna of alate form. I. — Geoica phaseoli, antenna of apterous form. J. — Geoica radicicola, antenna of alate form. K.— Geoica lucifuga, antenna of alate form. L. — Paracletus cimiciformis, apterous form. M. — Paracletus cimiciformis, eye of apterous form. N. — Paracletus cimiciformis, intermediate eye. O. — Paracletus cimiciformis, median thoracic wax plate of alate form. P. — Paracletus cimiciformis, foot of alate form with, one claw. Q. — Paracletus cimiciformis, antenna of apterous form. R. — Paracletus cimiciformis, forewing. S. — Paracletus cimiciformis, antenna of alate form. T. — Uamamelistes spinosus, hibernating apterous form. U. — naviamelistes spinosus, -wings. V. — Hamamelistes spinosus, cauda and anal plate of alate form. W. — Ham/imelistes spinosus, antenna of alate form. X. — Hamamelistes spinosus, portion of same, more enlarged. Y. — Hormaphis hamaTnelidis, wings. Z. — Hormaphis hamamelidis, antenna of alate form. AA. — Hormaphis hamamelidis, antenna of stem mother. BB. — Hormaphis hamamelidis, apterous form. CC. — Hormaphis hamamelidis, antenna of apterous form. DD. — Hormaphis hamamelidis, cauda and anal plate of apterous form. TLF,.— Hormaphis hamamelidis, dorsal wax pores. FF. — Hormaphis hamamelidis, cauda and anal plate of alate form. 107 PLATE XV. A. — Oregma lanigera, apterous form. B.—Oregma lanigera, antenna of alate form. C. — Oregma lanigera, caucia and anal plate of alate form. D. — Oregma lanigera, wings. E. — Oregma lanigera, end of abdomen of apterous form. F. — Oregma lanigera, antenna of apterous form. G. — Oregma bambusae, head of apterous form. n. — Glyphinaphis bambusae, apterous form. I. — Glyphinaphis bambusae, antenna of apterous form. J. — Glyphinaphis bambusae, cornicle of apterous form. K. — Glyphinaphis bambusae, cauda and anal plate. L. — Mansakia miyabci, wings. M. — Ceratoglyphina bambusae, apterous form. N. — Ceratoglyphina bambusae, cauda and anal plate. 0. — Ceratoglyphina bambusae, cornicle of apterous form. P. — Ceratoglyphina bambusae, antenna of apterous form. Q. — Astegopteryx distychii, antenna of alate form. R. — Astegopteryx distychii, fore wing. S. — Astegopteryx distychii, anal plate of alate form. T. — Astegopteryx distychii, cauda of alate form. U. — Astegopteryx distychii, end of abdomen. V. — Astegopteryx styracophila, fore wing. W. — Astegopteryx gallarum, fore ^ving. X. — Astegopteryx gallarum, tip of antenna of alate form. 108 Bui. 826, U. S. Dept. of Agriculture. PLATE XV. Bui. 826, U. S. Dept. of Agriculture. Plate XVI. PLATE XVI. A. — Aleurodaphis blumeae, apterous form. B. — Aleurodaphis blumeae, antenna of apterous form. C. — Aleurodaphis blumeae, dorsal wax pores. D. — Aleurodaphis blumeae, margin. E. — Aleurodaphis blumeae, cauda and anal plate. F. — Cerataphis lataniae, apterous form. G. — Cerataphis lataniae, wings. 11. ^Cerataphis lataniae, cauda and anal plate of apterous form. I. — Cerataphis lataniae, antenna of apterous form. J. — Cerataphis lataniae, antenna of alate form. K. — Cerataphis lataniae, anal plate of alate form. L.-^ Cerataphis lataniae, cauda of alate form. M. — Cerataphis lataniae, head of apterous form, central view. N .—Thoracaphis arboris, apterous form. O. — Thoracaphis arboris, antenna of apterous form. P. — Thoracaphis arboris, cauda and anal plate of apterous form. Q. — Thoracaphis ficus, antenna of apterous form. R. — Thoracaphis ficus, extremity of abdomen, apterous form. S. — Thoracaphis castaneae, wings. T. — Thoracaphis castaneae, distal segment of antenna of alate form. U. — Thoracaphis castaneae, cauda and anal plate of alate form. V. — Thoracaphis castaneae, cornicle of alate form. W. — Thoracaj>his castaneae, antenna of alate form. 109 ADDITIONAL COPIES OF THIS PUBIJCATIGTi MAY BE rROCXniED FROM THE SUPEKrNTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. VT 39 cent:^ per copy jjiv^ir^SiC's:^ UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 833 Contribution from the Bureau of Entomology. L. O. HOWARD, Chief. J^^'^U Washington, D. C. May 31, 1920 CHRYSANTHEMUM MIDGE.^ By C. A. Weigel and H. L. Sanford, Collaborators, Tropical and Subtropical Fruit Insect Investigations. CONTENTS. General description 1 Historical 2 Distribution in the United States and Canada. . 3 Varieties affected 4 Economic importance 4 Technical description 6 Page. Life history and habits 7 Natural enemies 13 Experiments in control 14 Summary of control and recommendations 21 Preventive measures 22 Literature cited 23 GENERAL DESCRIPTION. The chrysanthemum midge, Diarthronomyia Tiypogaea (F. Low), is a European insect which gamed entrance into the United States a few years ago and smce then has been reported as injurious from widely separated localities in this country and Canada. When chrysanthemums are infested by this midge, the attention of the casual observer is most likely to be drawn to the presence of galls. These galls occur on the leaf, stem, or flower head of the chrysanthemum plant. (See PL I, B and C.) After the larvse hatch from the orange-colored eggs, which are deposited by the adult female on the surface of tender tips and new growth, they bore their way into the tissues, thereby givmg rise to the galls. The galls are cone-shaped and generally project obliquely from the surface. The length of the gall when fully developed is about one- 1 The account of the chrysanthemum midge contained in this bulletin is the result of an investigation which was undertaken with the intention of providing fiurther data on the life history and habits of this insect, as well as more satisfactory means of control. The preliminary life-history studies which were started in February, 1917, by A. D. Borden, of the Bureau of Entomology, were subsequently taken up by H. F. Dietz, of the Federal Horticultiu-al Board. Further studies on life history and control were inaugurated and carried to completion by the writers as a result of the many inquiries for advice as well as the outbreaks reported during the chrysanthemum season of 1918 and the spring of 1919. During the entire period the work has been under the constant supervision of E . R, Sasscer, chief inspector of the Federal Horticultural Board and collaborator of th« Bureau of Entomology. 150054°— 2a-Bull. 833 1 2 BULLETIN 833, V. S. DEPARTMENT OF AGRICULTURE. twelfth of an inch. When the leaf is affected, the galls usually occur on the upper surface. In such cases a slight swelling often may be observed on the opposite or under side of the leaf. Growth and development of both larva and pupa take place within this gall. When the pupa is fully developed it pushes itself out of the gall, still inclosed in the pupal skin. The latter then splits down the middle of the head and back to allow the adult to emerge. (See PI. II, C.) In the adult stage the midge is a fragile two-winged fly, one- fourteenth inch in length. The abdomen of the male is yellowish orange, while that of the female is reddish orange. The adult on emerging leaves its pupal skin protruding from the opening of the empty gall. (See PI. II, A.) As shown in the life-history studies, the adults emerge after midnight and egg-laying takes place early in the morning. HISTORICAL. In 1870, E. Perris (1, p. 177)i is recorded as having observed ceci- domyiid larvae on Leucanthemum vulgare. Six years later J. E. Von Bergenstamm and P. Low (2) also recorded larvae which in all prob- ability were cecidomyiid, as they were found attacking the young leaves of Chrysanthemum leucanthemum {Leucanthemum vulgare.) In August, 1875, E. Berroyer collected specimens m Raxalpa, a group of the Eastern Alps of Austria-Hungary, at an elevation of 5,000 feet above sea level. These specimens, which consisted of subterranean galls, and two adult male midges, were submitted to Franz Low (3) by Von Bergenstamm. Franz Low made the original description in 1885, which freely translated is as follows: Male: Antenna 2-14 jointed. Peduncle ehorter than tlie segments. Terminal segment with two whorls of stout setse. Wings cloudy white. All veins, also costal vein white. Second parallel vein straight, disappearing in the apex of the wings. First branch of the third parallel vein so weak that it is not visible, unless under best of light and high power, but the wing fold on the other hand is very plainly discernible. Halteres white. Legs with closely adhering hairs, also appearing white. Larva: The still unknown larva causes galls on the underground parts of the stem of Chrysanthemum atratum Jacq., in which their complete transformation takes place. In each are also found several larvae, each larva occupying one cell for itself. Pupa: The pupa has (like the pupa of the Asphondylia species) three pairs of horns, but on the other hand, lacks the rows of delicate horns foimd on the dorsal surface of abdominal segments of the Asphondylia species. The two cephalic horns are very large, widely separated, slender, quite pointed and with the point somewhat recurved. The horns of both other pairs are very small, alike in size and structure, pointed and also with the point somewhat recurved . They occur in such a manner that a horn is located above and below each eye of the pupa; consequently, the horns of one pair are widely separated from each other. In the cells of the galls in which pupation of the larva takts place, the pupae are so situated that the head points to the periphery of the gall. 1 Numbers in parentheses refer to " Literature cited," p. 23. CHRYSANTHEMUM MIDGE. Galls: They occur, three or four in number, at the extremity of the upper and lower surfaces of the stem of the Chrysanthemum atratum Jacq., which is irregularly rounded, varying in size from a large hemp seed to a large pea. Outside naked, consisting of a fleshy homogeneous mass in which occur small, elongated cells or chambers each bein"- inhabited by a larva. Distribution: On Raxalpa about 5,000 feet above sea .evel. (E. Berroyer.) From this date, until it was first reported by Felt in the United States, frequent references are found in literature relating to its occur- rence and synonymy. Among these may be hsted the followmg: Riibsaamen (4, p. 375),Kieffer (5, p. 21), (6), (7, p. 351, 353), (13), Baldrati (8, p. 40-41), Kertesz (9, p. 69), Houard (11), and Kiister (12, p. 77, 274). On March 26, 1915, R. H. Pettit of the Mchigan Agricultural Col- lege had his attention called to the existence of an infestation occur- ring in large chrysanthemum houses at Adrian, Mich. vSpecimens were submitted to E. P. Felt (14), State ento- mologist of New York, who in April of that year determined it as the chrysanthemum midge, Diarthronom- yia Jiypogaea (F. Low) . In 191 6 A. Gibson (21) reported the occur- rence of the midge at Ottawa, Canada. Subsequently, reports of its occurrence have been received from widely separated points, both in the United States and Canada. According to California florists, this pest has been present in that State for about 15 years. DISTRIBUTION IN THE UNITED STATES AND CANADA. To date, the midge has been reported from California, Connecticut, Delaware, the District of Columbia, Georgia, Illinois, Indiana, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Hampshire, New Jersey, New York, Ohio, Oregon, Pennsylvania, Rhode Island, South Dakota, Tennessee, and Virginia, as well as from Ottawa and Victoria, in Canada. Since the first records of its occurrence and seriousness were report- ed from the Middle West it would appear that the distribution of infested plants from this region was the source of its further spread to the other sections of the United States and Canada. Its distribution, as indicated on the accompanying map (fig.l), is almost wholly confined to the larger chrysanthemum-growing regions of the United States, which include the areas surrounding the Great Fig. 1. — Map showing distribution of the chrysanthemum midge in the United States and Canada. 4 BULLETIN 833, U. S. DEPAKTMENT OF AGRICULTURE. Lakes and the northern and central Atlantic seaboard. On the Pa- cific coast it is reported from California and Oregon, and its presence has been confirmed in South Dakota, Tennessee, and Georgia. Doubtless a careful survey would locate many additional isolated infestations which have not been brought to the attention of State or Federal entomologists. VARIETIES AFFECTED. While the chrysanthemum midge has been recorded from central and southern Europe as seriously injuring the comm.on white or ox- eye daisy (Ohrysanthemum leucantJiemum) , as well as C. corymbosum, C. atratum, C. japonicum, and C. myconis (11), its depredations in North America are confined to practically all of the conomercial chrysanthemums, both the single and pompon varieties. The first infestation in this country was reported on the variety Mistletoe, and according to Felt (14, 15, 16) this variety appears to be very susceptible to the attacks of the midge. Although several attempts have been made to infest the Shasta daisy and the common field daisy, C. leucanthemum, it has not been possible to get the ovipositing female to lay eggs on them. This is of much importance, for should the infestation spread to this common weed, there would be great difficulty in eradicating this pest. A. Gibson lists the following varieties as being fairly free from injury : Bob Pulling, Gertrude Peers, Daily Mail, Oconta, Mrs. G. C. Kelly, W. Wood Mason, F. T. Quilleton, and E. T. Quittington. All the above varieties are the blended product of C. indicum and C. mori- folium, both of which grow wild in China and Japan. He reports the following varieties as being practically ruined: Smith's Advance, Ivory, Bonnaffon, Wm. Turner, Western King, and Englehart. Observations made by H. F. Dietz in the Middle West showed that the Wm. Turner variety had lost all the crown buds. In case of a thick infestation on the variety Dr. Enguehard all of the plants had to be discarded early in the season; also all Chadwick varieties, as well as Elberon, Major Bonnaff on, and Golden Mensa. The varie- ties which seemed to be least injured were Golden Age, Harvard, and White Bonnaffon. In some of the greenhouses of the District of Columbia during the season of 1918-1919 such varieties as Mensa, William Turner, and the white and yellow Bonnaffon were completely ruined on account of the severe infestation. ECONOMIC IMPORTANCE. Although a comparatively recent introduction this insect now seems to be firmly established in the United States and is one of the most important pests to be reckoned with by chrysanthemum growers. Bui. 833, U. S. Dept. of Agriculture. Plate I, The Chrysanthemum Midge. A, Adults emerging. JS, galls of midgo on chrysanthemum leaf. C, injury to terminal growth. I>, adult emerging from stem gall. 5ul. 833, U. S. Dept. of Agriculture. Plate 1 1 . The Chrysanthemum Midge. A, Empty pupal skins protruding from galls after emergence of adults. B, adults killed while emerging as a result of being sprayed with nicotine sulphate. C, adult female midge. CHRYSANTHEMUM MIDGE. 5 While it is primarily a greenhouse pest, A. Gibson, in August, 1915, found it occurring on both greenhouse and outdoor plants at Ottawa, Canada. Well-developed galls on the newly unfolded leaves were found by the authors, March 26, 1919, on hardy chry- santhemums which had been grown out of doors all winter at the Arlington Farm, Rosslyn, Va. A careful inspection of the entire stock revealed the fact that the insects had wintered over on these plants in the immature stages, probably as either larvse or pupss within the galls. Inasmuch as there were many empty galls present on the old dead and dried leaves of the previous season's growth, it was evident that this was an infestation of long standing. On April 11, many adults were found entangled in webbing spun by spiders among the developing new tip growth. These emerged from the above mentioned galls. The first severe infestation of the midge brought to the attention of the Bureau of Entomology was on chrysanthemums in greenliouses at Philadelphia, April, 1917. The entire stock was infested, causing a total loss to the grower. During the same year other florists reported a total loss of their stock of chrysanthemums valued at several thousand dollars. Even in the case of a light infestation the foliage is practically valueless for commercial purposes, and in the case of a heavy infestation the growth is completely arrested, (See PI. I, C.) Owing to numerous reports of injury by the chrysanthemum midge in the States of Indiana, lUmois, and Michigan, H. F. Dietz con- ducted a survey in this territory during the months of November and December, 1918, for the purpose of ascertaining the exact amount of damage occasioned by this pest. From the data gathered, this general locality proved to be one of the centers from which the pest was being distributed over the United States. Greenhouses of 33 florists were visited, including some of the largest growers and dis- tributors in this country. It was evident that if serious pests were estabhshed in this region, it would be only a matter of a year or two before such insects would be widely scattered. This is exactly what has happened in the case of the chrysanthemum midge. This pest was found established in 8 of 33 places visited. Serious injury to infested plants was noted. For example, no plants of the variety Dr. Enguehard came into flower, on account of their dwarfed, knotted, and knarled condition, with the result that the new central stem did not form. Moreover, other varieties, including Wm. Turner, were attacked just at the time the crown buds were "setting,' ' causing these flowers to become distorted. As a result, the flowers are not borne upright as normal flowers should be. Such evidence may be taken as an indication of the presence of the midge, even though the galls may not be numerous enough to attract at- tention. 6 BULLETIN 833, V. S. DEPARTMENT OF AGRICULTURE. From these data it is apparent that the insect is a serious pest, especially in cases where growers require from 50,000 to 100,000 plants for their own use and five to ten times this number to fill annua] orders for shipment. JP.2: j'nc.^jri--! ^'s Fig. 2.— The chrysanthemum midge: A, Leaf covered with galls; B, a single gall, more enlarged; C, gall cut open from above, showing young larva within; D, two galls cut in vertical section; E, larva, enlarged about 13 times; F, pupa; G, fly emerging from a gall on right, discarded pupal skin remaining in opening ofgallonleft; F, adult female fly, enlarged about 13 times; /, eggs. (9th Ann. Kept. State Ent. Ind.) TECHNICAL DESCRIPTION. ' ^gg [fig. 2, I]. — Reddish orange, length .15 mm., diameter .03 mm., the extremities narrowly rounded. Gall [fig. 2, A, B, C, D, GJ. — Conical swelling projecting obliquely from the surface. Size about 2 mm. in length. Color, generally green. Reddish green when growing outside and depending on variety affected. Tips appear dried out and grayish at ' Technical description of egg, larva, pupa, male, and female taken from Felt (20). CHRYSANTHEMUM MIDGE. 7 times. Hairs on tip in superabundance. Occurring on both leaf surfaces, stem and flower head. Larva [fig. 2, 0, E]. — Length 1 mm., yellowish or yellowish orange when full grown, moderately stout, the extremities rounded ; segmentation distinct and the skin smooth. Pupa [fig. 2, F]. — Length 1.25 mm., stoiit, narrowly oval, the cephalic horns dis- tinct, conical, the thorax yellowish orange, the wing pads fuscous in pupse nearly ready to transform, the leg cases dark yellowish brown, the abdomen a variable orange, narrowly roimded apically. Male. — Length 1.75 mm. Antennae nearly as long as the body, sparsely haired, fuscous yellowish; 17 or 18 segments, the fifth with a stem about three-fourths the length of the subcylindric basal enlargement, which latter has a length about twice its diameter and a rather thick subbasal whorl of long, stout setae; terminal segment variable, usually somewhat reduced, irregular, elongate, ovate. Palpi: The first segment subquadrate, the second narrowly oval. Mesonotum dark brown, the sub- median lines yellowish. Scutellum and postscutellum fuscous yellowish, the abdo- men mostly a pale yellowish orange. Wings hyaline, costa light straw, halteres yellowish transparent. Legs a pale straw, the pulvilli a little longer than the long, slender claws, the latter with a long, slender tooth basally. Genitalia; basal clasp segment moderately long, stout; terminal clasp segment short, stout, with a distinct spur; dorsal plate short deeply and roundly emarginate, the lobes short, broad, obliquely truncate apically; ventral plate short, deeply and roundly emarginate, the lobes rather long and tapering to a narrowly rounded apex. Female [fig. 2, H]. — Length 1.75 mm. Antennte extending to the third abdominal segment, sparsely haired, fuscous yellowish; 16 or 17 segments, the fifth with a stem about one-third the length of the cylindric basal enlargement, which latter has a length a little over twice its diameter; terminal segment reduced, sometimes com- pound and tapering to a narrowly rounded apex. Palpi: The first segment subquad- rate, the second subconical and with a length a little greater than the first. Mesono- tum fuscous brown, the submedian lines, the posterior median area, the scutellum and postscutellum mostly fuscous yellowish, the apex of the scutellum narrowly fuscous. Abdomen reddish orange, apically fuscous yellowdsh, the o^dpositor about one-half the length of the body; terminal lobes short, broad, broadly rounded and sparsely setose apically. Other characters practically as in the male. LIFE. fflSTORY AND HABITS. ADULT STAGE. The adult female shows a marked preference to lay eggs in the buds, or in the tissues just unfolding from the buds. Serious injury to the host results from this habit, and the commercial value of the plant is greatly reduced, if not entirely eliminated. Emergence of adult. — On March 26, 1919, H. L. Sanford observed the emergence of a female at 1.09 a. m. On the following night observations by C. A. Weigel taken at intervals varying from 15 to 20 minutes apart beginning at 12.30 a. m., and ending at 5.45 a. m., showed the emergences as follows : At 1 a. m., 1 female; at 4 a. m., 1 male; at 4.02 a. m., 4 females, 3 males. In addition the following observations were recorded: When the female emerges the body is pushed out of the gall for approximately three-fourths its length. The wings at this time appear as dark gray or black club-like append- ages. The antennge are moved about in a very active manner. The legs, which are at first folded parallel to the body, are thrust outward 8 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. and immediately begin movement. The body is swayed back and forth with contractions of the abdomen v/hich is still inserted within the gall, the latter acting as a fulcrum. In about two minutes from the time at which the adult is first seen to break through the gall, the entire body is freed from its pupal skin. The pupal skin remains protruding from the empty gall case, as seen in Plate II, A. With a convulsive motion the insect gains a foothold by thrusting out the legs and was observed almost immediately to travel rapidly to the underside of the leaf, assuming a position at right angles to the leaf sm-face. In this hanging position, which is characteristic of the adult female, it remained over an hour, occasionally fluttering its wings. Upon being disturbed it made a very short flight and re- turned to the underside of the same leaf. Evidently the above- described position is assumed to allow for drying and inflation of the wings. Mating. — A male was observed to fertilize a female approximately 3 minutes after the emergence of the latter, the entire operation taking 10 seconds. One male was observed to fertilize three females in rapid succession. Table I. — Longevity of adults and date of ovi position. Adult emerged. 1917. Apr. 2, a. m.. Apr. 5, a. m. . Apr. 6,a. m. . Apr. 10, a. m . do Apr. 11, a. m. Apr. 12, a. m. Apr. 13, a. m. .do. do Apr. 17, a. m. 1918. Jan. 31, a. m.. do do do Feb. 1, a. m . . Feb. 11, a. m. Mar. 14, a. m. July 10, a. m. July 17, a. m. July 18, a. m. July 19, a. m. Aug. 2, a. m.. Aug. 3, a. m. . Aug. 5, a. m.. Sept. 2, a. m.. Number. Male Fe- male, Oviposition. .\dul(s died. None... do.. do.. Apr. 10. None... .do. .do. Apr. 13. .do. None.. ....do. ....do.. Jan. 31 . do.. do., Feb. 1.. Feb. 11. July 10. Aug. 3. Aug. 5 . Male. Apr. 6, p. m.,2. Apr. 7, a. m., 1. Apr. 10, p. m Apr. 10, p. m. Apr. 13, p. m., 2 Apr. 14, a. m. Apr. 13, p. m. Lost. Mar. 15, a. m. July 17, p. m. July 20, a. m . Aug. 3, a. m.. Sept. 2, p. m. Female. Apr. 5, a. m. Apr. 8, a. m. Apr. 11, a. m. Apr. 12, a. m. Apr. 13, p. m. do Apr. 14, a. m. Apr. 14, a. m. Apr. 14, a. m. Apr. 17, a. m. Apr. 20, a. m. Feb. 1, a. m.. Feb. 2, a. m.. Feb. 1, a. m.. Feb. 2, a. m. . do Feb. 13, a. m. July 11, a. m. July 18, a. m. July 20, a. m . Aug. 3, a. m. Aug. 6, a. m. Length of life. Days. 3 3 i I i 1 i »2 >2i H 2 i • 1 li i li ' 4 3 Female not fertilized and did not deposit eggs. CHRYSANTHEMUM MIDGE. 9 Longevity of adults. — In the life-history studies it was found that males usually live less than 12 hours, while the females live from 2 to 3 days, in captivity and when not mated. Mated females probably do not live over 24 hours. In the case of tliree males which were seen to emerge after 1 a. m,, all were found dead at 5.45 a. m. These re- sults are given in detail in Table I. Activity. — The males are not often seen on the wing, as they die shortly after mating, which takes place during the early morning hours, almost directly after emergence. When on the wing, how- ever, they are very active. The females, while generally rather sluggish, may show great activity at times, especially in the early morning and just prior to egg deposition. Dietz observed that on cool days when the temperature was 50° F. in the daytime, and cooler at night, the adult females were exceedingly active. Emergence rec- ords taken on many cages show that there are a somewhat larger nimiber of females than males, the actual count being 464 females to 365 males. Egg de'position. — The female when laying eggs keeps her legs at equal distances apart and the ovipositor held at right angles to the rest of the body, beginning probably at the third or fourth abdominal segment. If not disturbed she acts with great precision. On one occasion a female was carefully observed during two operations, laying 34 and 19 eggs respectively. The first ovipositiori period lasted exactly 3 minutes and 30 seconds, the second only 1 minute and 20 seconds. An interval of only 10 seconds elapsed between the oper- ations. The tip of the long flexible ovipositor is thrust between the pubescent hairs of the young terminal growth and pressed down firmly on the surface. During deposition the eggs may be seen moving down the tube with great rapidity, accompanied by slight contractions of the ovipositor. Because of its darker color, the egg is plainly visi- ble through the walls of the ovipositor. The number laid at one time, according to Table II on page 10, varies from 5 to 135, with an aver- age of 32. Guyton (31) found that "the number deposited by each female is from 80 to 150." The eggs are usually placed in irregular masses, although they may be placed in strings or chains. During the whole operation the ovipositor is thrust around in a very nervous manner. At times when the female moves slightly about, the body is not straightened out but dragged along to the new place of deposi- tion, in the original bent position. Although egg-laying may take place up to and later than 4.30 p. m., the maximum egg-laying occurs from the early morning hours up to midday. The midges prefer to lay eggs on the terminal growth of the plant on which they emerge, and particularly on the unfurling tip growth. This is true even in cases 150054°— 20— Bull. 833 2 10 BULLETIN 833, U. S. DEPAKTMENT OF AGRICULTURE. where the terminal growth has been badly distorted previously as a result of their work. Females have been noted to return to such plants from which they have emerged and deposit eggs thereon. Even though they were repeatedly distm-bed, they would return quickly and continue deposition. On many occasions it was found that when these 3^oung tips were separated and examined, the eggs would be present in greater numbers between the folds of the new tissue than on the neighboring exposed surfaces. Incubation. — Observations, as given in Table II, show that the in- cubation period varies from 3 to 16 days. The time of incubation may be affected by the season of the year. Table II. — Length, of egg -period. Host plant De- posited. ■Hatched. Length. Number deposited. October Frost Snow Queen Silver Wedding... Golden dlow Harvard Esanath Snow C^ueen MonroA ia Martha Saunders. Gold C'hadwick... Ksanath Monrovia Silver Wedding do do Dr. Enguehard Pompon Silver Wedding Dr. Enguehard Dr. Pool Mrs. R. N. McLuckie. M argaret Bissett Silver Wedding Mensa Silver Wedding Golden Wedding 1017 Apr. Apr. Apr. Apr. Apr. Apr. Apr. in Apr. 12 Apr. 13 Apr. 16 Apr. 17 Apr. 20 1918. /Jan. 31 \Feb. 1 ..do Feb. 2 Feb. 11 Feb. 28 Mar. 1 Mar. 4 Aug. 2i"i Aug. 27 Aug. 29 ..do ..do Sept. 5 Sept. 21 Apr. 7 Apr. 8 Apr. 9 Apr. 10.... Apr. 11.... Apr. 12.... Apr. 1.") Apr. 17.... Apr. IS.... Apr. 21.... Apr. 22.... Apr. 25.... JFeb. 9 Feb. 7 ....do Feb. 15-19. Mar. 6 Mar. 6-10.. Mar. 9 Sept. 6 Aug. 30. .. Sept. 4-14. Sept. 6.... Sept. 6-14. Sept. 10-14 Sept. 24... Days. 5 5 5 5 5 5 5 5 5 5 5 5 8-9 G 5 4-8 6 5-9 5 11 3 6-16 8 8-16 5-9 3 No records. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. 25 eggs. No records. Do. Do. Do. 34 eggs. 19 eggs. 5 eggs. 20 eggs. 24 eggs. 25 eggs. 135 eggs. Development. — The development of the larva within the egg can be made out easily. In the newly laid egg the nucleus, which is reddish in color, is rather central, but shows a slight tendency to be located toward the anterior end. There are approximately 18 yolk bodies present which are arranged in a row of 9 on each side. The row^s do not extend quite to the posterior end of the egg. As development proceeds the red pigment or nucleus moves to one side of the egg, which later is apparently the ventral surface. In the meantime the yolk bodies collect in a mass near the nucleus. At each end the small air spaces, which at first are relatively small, later become much in- creased in size. The segmentation then begins to appear, with a simultaneous contraction of the entire contents. A redistribution of CHRYSANTHEMUM MIDGE. 11 the yolk then takes place, an equal portion being present in each seg- ment. At this stage the pigment takes up a posterior position. Finally the complete outline of the larva is visible tlu-ough the shell. At this stage the darker head and mouth parts can be seen, and the previous yolk material appears as a chainlike formation which ex- tends from the anterior to the posterior end throughout the entire inner or central portion of the body. Distinct segmentation is now clearly visible. Hatching. — Borden observed the larva when hatching to break from the eggshell by a small cap at one end. Complete development of the larva and pupa takes place within the gall. LARVAL AND PUPAL STAGES. The larva or maggot upon hatching moves about on the surface among the plant hairs for a period of from 1 to 3 days, preparatory to boring into the tissue. It varies in color from a transparent white to pale orange when seen with the aid of a binocular just after boring beneath the epidermis. On February 2, 1918, H. F, Dietz observed larvae, which hatched from eggs laid four days previously, boring into tissue. (10 X eyepiece, 24 mm. objective, binocular.) The observations were as follows : One larva which was found half buried in the tissues of the stem was timed until it disappeared. It required 12 minutes for complete disappearance. During this period the larva moved back and forth with an irregular spiral movement, about 30 seconds being necessary for one complete movement back and forth. Several larvge were then observed beginning to bury themselves, but the operation was interrupted and discontinued, the interruption being caused by another larva which was slowly crawling about in search of a suitable place to "dig in. " One of these larvae was observed crawling slowly about for approximately 3 minutes. The bright red pigment of the posterior third of the body is very characteristic at this particular stage. As a result of the larva boring into the tissue, an irritation is pro- duced wliich results in the production of swellings or galls on the plant containing the developing larvae and pupae. No molts have been observed from the time of hatching to the time at which the larva is seen entering the tissue. The larva lies bathed in a fluid within the gall. The fully developed larva is of an orange color. The female pupa is usually orange colored with the head, thorax, legs, and wing pads nearly black, while the male may be of a lighter or straw color. Formation of the pupa takes place about two weeks from the time the egg is hatched. It is white at first with only a slight brown tinge about the head, but later the head, thorax, wing- 12 BULLETIN 833, U. S. DEPAKTMENT OF AGRICULTURE. pads, and legs are dark brown, and the abdomen orange. The cephalic horns are distinct in the nearly mature form of the pupa. When mature the pupa works its way out of the gall. On emer- gence a split is made upon the dorsal line of the head and thorax of the pupal case through which the adult issues. (PI. I, A, D.) During emergence the adult is very active and issues very rapidly. From the time that the larva enters the tissue to the first sign of a swelling, or gall, observations on 18 life-history cages show that a period ranging from 4 to 14 days elapsed with an average of 7 days (see fourth column of Table III). The young gall may now be readily recognized by the characteristic white spots, or slight swell- ings. It takes from 21 to 46 days, with an average of 28 days, from the time at which the larva first enters the tissue until the emergence of the adult. These observations were taken from the results obtained from 18 cages during the spring of 1917 and the spring and fall of 1918, as is shown in the following table. Britton (32) found it to require from 20 to 50 days to transform within the gall. Table III. — Time required from: (a) larva entering tissue until first sign of gall; (b) larva entering tissue until emergence of adult. Cage. Larva entering tissue. Apr. 7-n. do.... do.... do.... 1918. reb.7 do Mar. 6-12 Mar. 6-10 Mar. 12 do do do Aug. 31 Sept. 2 Sept. 4-14 Sept. 4-11 Sept. 7-14 Sept. 24-25.... First sign of gall. Feb. 1.3-lG.. do Mar. 15 do Mar. IS Mar. 16 Mar. 18 do Sept. 9-19.. Sept. 9 Sept. 17.... Sept. 11.... Sept. 14-19. Oct 2 Time, rec|uircd (a). Days. 7 7 6 7 6 M 6 6 2 14 7 8 7 Date of emergence. Apr. 29 to May 2. Apr. 30 May 1-2 Apr. 30 to May 2. Mar. 4 Mar. 4-7.. Apr. 6-22. Apr. 8-22. Apr. 27... do Apr. 6.... Sept. 25 Oct. 2 , Sept. 20 to Oct. 3. Sept. 27 to Oct. 2. Oct. 4 Oct. 24 to Nov. 2. Total length (b). Day^. 22 121 23 22 25 28 30 38 M6 24c, 25 25 30 121 22 23 34 ' Minimum time. - Maximum time. From Table IV it is evident that the total life cycle requires from 27 to 52 days, with an average of 35 days. These data represent the results obtained from 17 life-history cages under Washington conditions. There is a constant overlapping of broods when the greatest numbers are present, namely in the spring and fall of each year. The aestivation period has been foiuid to extend, in Maryland, Virginia, and the District of Columbia, from the early part of June to the latter part of August. CHRYSANTHEMUM MIDGE. Table IV. — Length of compUtc life eyc.le. 13 Cage. Egg.s depoijited. Adult emerged.' Length of life cycle. Apr. 2. do. do. Apr. 29 to May 2. Apr. SOtoMay 2. May 1 to May 2. . Feb. 1.. Feb. 2.. Mar. 1 . . Mar. 5 . . Mar. 0-9 Aug. 3.. do.. Aug. 0. . Aug. 27. Aug. 29. do.. do.. do.. Sept. 21. Mar. 4 Mar. 4 to Mar. 7 Apr. 8 to Apr. 22... Apr.G Apr. 22 to Apr. 27.. Sept. 4 to Sept. 0... Sept. 11 to Sept. IS. Sept. 5 Sept. 2.5 to Oct 4. .. Sept. 2(>toO''t3. .. Sept. 27 to Oct 2... Oct 2 Oct. 4 Oct 24 to Nov 2 Dai/s. 27-30 28-30 29-30 31 30-33 3S-52 32 47-49 32-34 39-40 30 29-38 28-35 29-34 34 30 33-42 I Data given in Table I .show that mated females deposit eggs on date of emergence. In the spring of 1917, 1918, and 1919, three distinct generations were observed. The first generation started about the middle of February and the last adults of this generation emerged during the last few days of April. The second generation started about the middle of March and the last adults issued around April 30. The third generation started the latter part of April and emerged during the early part of June. In the faU of 1918 when the occurrence increased again a similar grouping of generations was evident, the first beginning about the latter part of August, and the last adults emerging during the first days of October. A second generation started about the middl& and latter part of September, maturing the first days of November. The third generation was observed beginning about the middle of October and the last adults emerged about November 25. NATURAL ENEMIES. Felt (20), in speaking of the natural enemies of the midge, states that— It v?-as very likely brought to America without the normal quota of parasites and for a time at least it may prove to be a somewhat difficult insect to control, though it would seem as if the native parasites of our large and varied gall midge fauna might in time prey most successfully upon this midge. Essig (18) mentions: During the summer a large number of parasites were reared from infested plants and one species in particular did excellent work in the university greenhouse. The material was sent away for determination and a few observations made as follows: Amblymerus sp. This hymenopterous parasite has been described by Mr. A. A. Girault, through the kindness of Dr. L. 0. Howard, and a description is to appear elsewhere. The adults are black with yellow markings on the legs. The females vary from 1 mm. to 1.2 mm. in length, and the males are somewhat smaller. The 14 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. larvae live within the galls alongside the maggots of the gallfly, which they gradually consume. They remain within the galls until mature, when they emerge through small circular holes. This species is the most abundant during the summer months and all of the adults were reared during August, September, and October. In not a few cases as high as 80 per cent to 90 per cent of the maggots were destroyed. Tetrastichus sp. The generic determination of this insect was made by Mr. Harry S. Smith, superintendent of the State insectary, Sacramento, Calif. It is also a small black parasite, somewhat larger than the former and easily distinguished from it by the foiir-jointed tarsi and other characters. EXPERIMENTS IN CONTROL. At the outset it must be borne in mind that it is very important that control measures should be closely related to the propagation of the stock, and moreover from the practical and economical point of view these steps should be undertaken at a period in the propagation when the plants can be severely cut back, allowing sufficient time to overcome any injury which may result from such treatment. Inas- much as cuttings are propagated from the stock plants of the previous season, it is evident that the logical time to eradicate an infestation of long standing is immediately after the flowers have been removed. Naturally such precautions as would absolutely safeguard the grower against further spread and increase would have to be inaugurated. An experiment conducted in one of the commercial greenhouses in the District of Columbia in which the plants had been heavily infested the previous season was carried out as follows : These heavily infested stock plants were heeled in and pruned back, and the entire portion above ground was thoroughly dusted with a mixture of equal parts of tobacco dust and air-slaked lime as often as the new growth appeared. As a result of such treatment the new growth was kept practically clean from further infestation and the plants at the same time showed a stimulated growth. The checks which were run simul- taneously had all the new growth badly infested. Occasionally it may happen that a grower's previous season's stock was clean and new material or varieties are being received in the form of cuttings or young plants from localities where the insect is known to exist. There is a possibility of introducing the midge on such stock even though no definite signs of the insect are visible. In the spring when cuttings and young plants are being exchanged, an infestation, especially in the egg or very young larval stage, will defy even the most careful and zealous inspectors. Interception made of plants harboring the eggs only can easily be overlooked by virtue of their concealed location in the still unfurled leaf bud. A precautionary treatment should be given at the time when the plants or cuttings are taken or received and would prove of inesti- mable value in killing the eggs and immature stages. The question of controlling an infestation already in the well- developed gall stage as well as the adult stage must also be con- CHKYSANTHEMUM MIDGE. 15 sidered. The control of the adult stage need not be taken up in experi- mental manner, in view of the known fact that such fragile fhes can easily be controlled by light fumigation with hydrocyanic-acid gas or by burning tobacco papers, provided it is done at the correct time. The preceding paragraphs clearly point out the advisability of determining the practical value of the several premaseg stated. Incidentally such factors must necessarily be solved independently of each other, but with the final view of either applying each separately when the case so warrants or consolidating such phases as would be most practical and consistent with commercial practices. The practical conclusions and recommendations that follow were deduced from results that are outlined and stated in the following pages. EXPERIMENT 1.— CONTROL OP EGGS BY MEANS QF DIPPING INFESTED CUTTINGS. Emphasis has been placed on the importance of beginning with clean cuttings in the spring. It, therefore, was found advisable to test out this point, and, if effective, such practices would be a safe- guard to growers whose stock was badly infested as well as growers who had received outside material. Three lots of six cuttings each were treated by J. L. Dietz as given in the following table. After treatment they were planted in sand. The tray, together with the plants, was then protected from further attacks by placing it under a close-mesh screen. The plants were treated September 26, and observations taken October 5 and 10 and November 1 are given in Table V. Table V. — Control of eggs by means of dipping infested cuttings. Ex- peri- ment. Treatment. Observations and results. 1 2 3 40 per cent nicotine sulphate (1-500); laundry soap 1 ounce to 1 gallon; tips of cuttings dipped. Sept. 26, 1918. Same treatment as above, except that entire cuttings were dipped. Sept 26, 1918. Check; no treatment Nov. 1: Eggs did not hatch; no new galls devel- oped; 4 plants well rooted, 2 plants poorly rooted. Nov. 1: Eggs did not hatch; no new galls devel- oped, neither did original galls develop any further; 4 plants well rooted, 2 plants poorly rooted. Nov. 1: Eggs hatched; 1 plant had 3 well-devel- oped galls present; 3 plants rotted off; 3 plants were well rooted. DISCUSSION OF RESULTS. Comparing the results of both treated lots with those of the check, the results are fairly conclusive. On the treated plants the eggs in each case did not hatch, neither did the young galls originally present make further progress, whereas, in the case of the check, while only one plant showed definite results, it is clear that the galls developed. The effect of such treatment on the plants is somewhat inconclu- sive due to the rotting off of three plants in the check. In the case of the treated plants, four survived in each lot. 16 BULLETIN 833, XT. S. DEPARTMENT OF AGRICULTURE. EXPERIMENT 2.— CONTROL OF EGGS BY MEANS OF DIPPING ENTIRE PLANT. The object of this experiment was practically synonymous with that of experiment 1 except that entire plants were used instead of cuttings as was the case in the previous test. Twelve plants on which newly laid eggs were present were divided into three lots, (a), (b), and (c), of four plants each. They were then treated as follows: (a) Four plants dipped in 40 per cent nicotine sulphate (1-500) plus laundry soap 1 ounce to 1 gallon of solution. (b) Foiu" plants dipped in 40 per cent nicotine sulphate (1-800) plus laundry soap 1 ounce to 1 gallon of solution. (c) Four plants as check; no treatment. In each case the plants were constantly protected by covering them with an ordinary lantern globe, the free end of which was screened with a double layer of cheesecloth. The first dipping was done on March 29, while subsequent dippings were done on March 31 and April 1,3, and 8. The plan was to treat them at least every 2 or 3 days for a week or 10 days. Observations were taken frequently on the effect of the treatment both on the plants and on the eggs. These obser- vations were continued until the close of the experiment, April 30, at which time the plants were uncovered and placed in the open green- house. Results. — (a) Effect on plants: Slight burning was encountered with the 1-500 strength solution, but with the 1-800 solution no such trouble was evident. The check plants showed some sooty fungus which was probably due to the presence of honeydew secreted by aphids. (b) Effect on eggs : During the first few days the eggs showed normal development within the shell on the treated plants, but larvae failed to hatch from them, and no galls developed. On the check plants all of the eggs hatched successfully and many galls developed to maturity. Conclusions. — -The expectations were fully confirmed by the re- sults obtained, as was the case in experiment 1. The checks devel- oped healthy and normal galls while in the treated plants further development was promptly arrested. From the economical point of view it is safe to say that nicotine sulphate (1-800) plus the soap will control the egg stage effectively if properly applied. TREATMENT OF BOTH CLEAN AND INFESTED CUTTINGS. The importance of having all new stock free and clean from the immature stages of the midge and of having the control at all times closely related to the propagation of stock has already been referred to. The above points were tested by the following experiment. This test consisted of tliree parts. Lots 1 and 2 were clean cut- tings, whUe lot 3 consisted of infested cuttings. Each lot was sub- CHRYSANTHEMUM MIDGE. 17 divided into three divisions or rows, namely, (a), (b), and (c). The last, (c), in each case served as a check on (a) and (b), which were the treated rows. The first treatment consisted of dipping the cuttings directly after they were taken, and the subsequent treatments con- sisted in spraying daily for a period of 7 days thereafter with a solu- tion of the same strength. Lot 1 consisting of clean cuttings to be used as a check was kept in a separate and uninfested unit for the purpose of determining the effect of such treatment on the cuttings. Lot 2 consisting of clean cuttings and lot 3 of infested cuttings were both tightly screened and placed in a propagating frame in an infested unit. In this manner it was hoped to determine the effectiveness of such treatment when both clean and infested cuttings had to be grown together. LOT 1. Six clean cuttings in each row. Row (a). Nicotine sulphate (1-800) plus fish-oil soap 1 ounce to 1 gallon of solution. Row (b). Volatile nicotine sulphate (1-800) plus fish-oil soap 1 ounce to 1 gal- lon of solution. Row (c). Checks; not treated. Results. — During the period of treatment it appeared as though all the plants were affected slightly. Row (a) : The lower leaves of all the cuttings turned yellow, but 10 days after treatment was discon- tinued all but one 'cutting were in fine condition. Row (b): Even though no direct signs of burning were evident, the plants appeared sickly during the second and third treatments. Ten days after the discontinuance of the treatments the cuttings were in better condi- tion than either (a) or (c). The check row (c) was slightly affected due to its proximity to the treated rows. Ten days after treat- ment was stopped only two plants were in good condition and one was poor. The conclusion to be draw^n from this test is that the method of treatment under (b) was probably the better of the tVv^o. LOT 2. Seven clean cuttings in each row. Placed in infested unit after first treatment. Row (a). Nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 ounce to 1 gallon of solution. Row (b). Volatile nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 ounce to 1 gallon of solution. Row (c). Checks; no treatment. Results. — Although this dosage was weaker than that used in lot 1, it affected the cuttings much more. After the second and third treatments the plants all looked wilted. Ten days after the last 18 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. application all the cuttings but one in row (a) were dead. In row (b) all looked very poorly. In (c), the check row, one cutting was dead while all the remainder looked fairly healthy. Conclusions. — It appears that the variety may have played an important part in this case. In this lot, as was so evident in the previous treatment, (a) causes injury, while (b) gives a double bene- fit— first, it does not cause injury; and second, it protects the plant. The midges spread to the check row (c), and two galls had adults emerge from them. LOT 3. Three rows of seven infested cuttings each. Row (a). Nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 ounce to 1 gallon of solution. Row (b). Volatile nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 ounce to 1 gallon of solution. Row (c). Check; not treated. Results. — This variety stood up much better than did that in lot 2. Row (a). Slight burning evident. In fair condition 10 days after treatment. Row (b). Only two poor plants; no burning. All recovered 10 days after treatment. Row (c). All plants in good condition, slightly better than row (b). No galls developed. Either treatment (a) or (b) might be used successfully. yiNAL CONCLUSIONS. Comparing the three lots with one another the following conclu- sions may be drawn: The practices described injure the cuttings only slightly and afford reasonable protection from the midges. EXPERIMENTS ON CONTROL OF GALL STAGE. The final consideration concerns the control of the gall stage, which is the hardest to combat. Owing to the habit of the larvae of burrowing mto the plant tissues, the chief difficulty is that most spraying mixtures when applied to foliage fail to exert their toxic properties against the insects contained therein. This may be accounted for by the resistance of the leaf structures to penetration by insecticides. EXPERIMENT 1. A preliminary experiment, in which was used 1 part 40 per cent nico- tine sulphate to 250, 500, and 1,000 parts water, respectively, to which fish-oil soap or laundry soap at rate of 1 ounce to each gallon of solution was added, showed conclusively that one application was entirely ineffective in controlling the gall stage. The data which were taken at various intervals showed that the adults emerged CHRYSANTHEMUM MIDGE. 19 successfully and continually. Hence, it was decided to test out the efficiency of several applications when made two and three days apart. The test is given in the next paragraph. EXPERIMENT 2. In the treated lots five plants were used for (a), (b), (e), and (f), and six plants for (c) and (d) for each test. The plants used had been growing in pots. Only two plants were used in both checks. The dipping was done as indicated every two or three days. Obser- vations were taken daily. Table VI. — Results of experiment 2 . Percent- Started. Closed. Treatment. Total emerged. Total killed. age of effi- ciency. Remarks and conclusions. 1919. (a)May 1... May 10.. Nicotine sulphate (1-500), flsh-oil soap 1 ounce to 1 gallon. 217 123 56.6 Greatest efficiency with- in 24 hours of treat- ment. (b) May 1... May 10.. Nicotine sulphate (1-800), flsh-oil soap 1 ounce to 1 gallon. Nicotine sulphate (1-1,000), 152 86 56.5 Same as above. (c) May 19. . May 23.. 45 30 66.6 Lower leaves turned yel- flsh-oil soap 1 ounce to 1 gallon. Volatile nicotine sulphate low and died on most plants. Slight blackening of tips (d) May 19.. May 23.. 72 35 48.6 (1-1,000), flsh-oil soap 1 on larger leaves. ounce to 1 gallon. (e)May 1... May 10.. Linseed-oil emulsion plus nicotine sulphate(l-800). 80 2 2.5 Not efficient. (f)May7.... May 9. . . Fish-oil soap 1 ounce to 1 gallon. Many. 0 0 All emerged successfully; not efficient. (g) May 1 . . . May 10. . Check Many. 0 0 All emerged successfully. (h) May 19. . May 23.. Check Many. 0 0 Same as above. Discussion of results. — The data represented in Table VI are self-explanatory as to the relative value of the various strengths and combinations of insecticides. The mathematical representation of the first two lots does not clearly define the actual state of affairs. It was repeatedly observed that on the day following treatment the number of adults killed was usually larger than on the second or third day. In other words, the effectiveness of such practices de- pends entirely on the interval which elapses between applications, and the conclusion to be drawn is that nightly or daily applications are absolutely necessary to get the maximum killing. The insecti- cide proved to be effective in killing the adult in the act of emer- gence, but did not sufficiently penetrate the tissues of the leaves to kiU the immature stages .within the gaUs. The column "Total kUled" is understood to mean the killing of the adult in the process of emergence (PI. II, B). It is also important that the treatment should be appHed when the adults are almost ready to emerge from the gaUs. 20 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. EXPERIMENT 3. In the foregoing experiment it will be observed that good results were obtained ^vith the nicotine sulphate diluted to 1-800. Another point upon which it seemed desirable to have more definite information was the effectiveness of applications made daily compared with those made two days apart. In the following experiment two lots were treated, while the third lot served as a check. The empty gaUs were pimctured and aU old pupal skins were removed from the plants in question. Eighteen plants were employed in each of the two treated lots, which were sprayed a total of five times. Five plants served as a check. Table VII. — Results on first and second days after treatment. Lot. Treatment. Results first day after treatment. Males, Fe- males. Total killed. Pupal skins. Results second day after treatment. Males. Fe- males Total killed Pupal skins. Nicotine sulpnate 40 per cent (1-800) and soap Volatile nicotine sulphate 40 per cent (1-800) and soap Checks 38 Conclusions. — The results are very conclusive and show distinctly that the best results are obtained by daily application. There is a remarkable decrease in the numbers caught on the second day fol- lowing treatment. Nicotine sulphate, therefore, used at the rate of 1-800 plus soap (1 oimce to 1 gallon) will control the adult on emer- gence, and hence is very well adapted to such cases where fumigation can not be followed consistently. EXPERIMENTAL WORK IN COMMERCIAL CHRYSANTHEMUM GREENHOUSES. Experimental work in a commercial greenhouse in Baltimore during the summer of 1918 seems to indicate that the chrysanthemum midge can be held well under control by sprajdng the infested plants with 40 per cent nicotine sulphate applied at the rate of 1 part of nicotine sulphate to 500 parts of water, with the addition of one- half omice of soap to each gallon of solution. No appreciable injury followed the application, although the plants were sprayed every other day for a period of six weeks. The application of this treat- ment by the grower resulted in his producing especially fine chrys- anthemums. It would appear that a double benefit was realized by such practices. In the fu'st place it controlled the insect, and, secondly, a distiuct stimulation of growth followed. This seems to confirm the work of Gossard (30) and Guy ton (31), who claim that 40 per cent nicotine sulphate dOuted with 500 parts of water and CHRYSANTHEMUM MIDGE. 21 fish-oil soap will kill the adult almost immediately after emergence, if the spray is applied not more than three or four days previously. A separate sash house was erected closely adjoining the greenhouse in which the spraying was done. Here 100 plants were fumigated nightly for approximately eight weeks, tobacco paper at the rate of 1 sheet to every 650 cubic feet of space being used. Wliile some burning resulted, the midge was held in check. From these two experiments it is evident that either practice would be efficient. A large commercial grower in Indiana reports that following in- structions of this bureau he entirely eliminated the midge from his houses by fumigating with tobacco papers every night from December 20, 1917, to March 20, 1918. Experiments conducted in several greenhouses in the District of Columbia in 1919 have indicated that consistent nightly fumigation is very effective in keeping down this insect. In several places where this was not advisable an effective spray was applied, consisting of nicotine sulphate (1-800) plus soap (one-half ounce to 1 ounce to each gallon of solution) . SUMMARY OF CONTROL AND RECOMMENDATIONS. From the life history, as well as from the experimental data thus far submitted, it is clear that certain points must be kept in mind if the best practical results are to be secured. First, several genera- tions are always present in greenhouses during the spring and fall occurrences ; second, the adults emerge and mate during the very early morning hours, and egg laying quickly follows; third, preliminary control experiments show that the egg stage may be controlled by means of spraying or dipping the cuttings or plants; fourth, it has been demonstrated that the adult can be killed easily at the time of emergence by consistent spraying; fifth, fumigation experiments in a commercial house proved that the adult is easily killed by fumi- gation either with nicotine papers or hydrocyanic-acid gas; sixth, experiments applicable to general propagation practices show con- clusively that such measures offer a reasonable safeguard and pro- tection against doubtful stock and infested material without injury to the plants. By adherence to a definite control program, involving any of the above cited measures, either singly or in combination, the insect can be readily controlled. In case of a very light infestation daily picking of gall-infested leaves will hold the pest in check. Should this practice prove ineffec- tive, nightly fumigation for a period of two or three weeks may be resorted to. 22 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. When a severe infestation is encountered the most heavily infested j plants should be taken out immediately and burned. This should then be followed by either fumigation or spraying as outhned below. Fumigate every night, with either nicotine papers or hydrocyanic- acid gas, for a period of at least six weeks. Tliis will kill all the adults that emerge during such a period and at the same time will prevent the fm'ther laying of eggs for future generations. The! dosage need not be very heavy in either case. When nicotine papers are used one sheet to every 1,000 cubic feet of space will suffice. If hydrocyanic-acid gas ^ is employed, one-eighth to one-fourth ounce per 1,000 cubic feet will kill all of the adults. The use of hydro- cyanic-acid gas is not recommended unless in the hands of a competent fumigator, owing to its deadly poisonous effects. Too much emphasis can not be laid on the fact that the fumigation must be set off after 12 o'clock, midnight, to be effective. It is preferable to start the generation between the hours of 12.30 a. m. and 2 a. m. Any fumi- gation done before midnight would be useless for it has been pointed out that the adult does not emerge until after midnight. On the other hand, if it is started later than 2 a. m. many adults will have emerged and laid their eggs. In case fumigation is not advisable, especially where chrysanthemum plants are isolated or when other varieties of plants are present in the houses, spraying is recommended. This must be done consistently for a period of fom- to six weeks, a 40 per cent solution of nicotine sul- phate extract diluted (1-800) being used, and soap added at the rate of one-half to 1 ounce per gallon of solution. The application should be made late in the afternoon in order that the best results may be obtained. In this manner practically all adults can be killed at the time of emergence and any eggs present will be destroyed. PREVENTIVE MEASURES. It has been proved that the means of disseminating this insect has been the shipment of infested chrysanthemums, both plants and cuttings. Interceptions made by the various State inspection officials, as well as by inspectors of the Federal Horticultm-al Board, definitely confirm this fact. It is therefore imperative that only clean plants and cuttings be brought into commercial houses where chrysanthemums are growing or to be growTi. Growers should care- fully examine all chrysanthemums received and see that all material intended for shipment or distribution is free from this pest. Any questionable material should at once be submitted to the State I For further information on the use of hydrocyanic-acid as in fumigating greenhouses, see Farmers' Bulletin 880. CHRYSANTHEMUM MIDGE. 23 experiment station or to the Bureau of Entomology, United States Department of Agriculture. The California Experiment Station (18) recommends the practice of growing the bulk of the chrj^santhemum crop under cloth as a satis- factory means of preventing the attack of the gall-fly. As a further preventive measure, it is recommended that cuttings be dipped in the following solution at the time they are taken: Nicotine sulphate, 40 per cent part. . 1 Water parts. . 800 Laundry soap ounce per gallon. . ^1 Another satisfactory method of securing chrysanthemum cuttings free from the midge when the previous season's stock has been infested is to plant the stocks in benches or cold frames directly after the season's crop has been removed. This should then be followed by thorough treatment with a mixture consisting of equal parts of dry or air-slaked lime and tobacco dust. It is advisable to keep all new growth covered with the mixture until further operations in the spring. LITERATURE CITED. (1) Perris, E. 1870. Histoire des insectes du pin maritime. (Diptferes). In Ann. Soc. Ent. Fr., V. 10, p. 135-232. Page 177: States that he observed cecidomyiid larva? in a gall formed by the union of the leaves of the central shoot of the terminal rosette of Lcucanthtmum tuJgarc. (2) Von Bergenstamm, J. E., and Low, P. 1876. Synopsis Cecidomyidarum. In Verh. zool.-bot. Gesell. in Wien, v. 26, p. 90, no. 516. Identical with (1). Reference taken from E. Perris. (3) Low, F. 1885. Beitrage zur NaturgescMchte der Gallenerzeugenden Cecidomylden. In Verh. zool.-bot. GeselL in Wien, v. 35, p. 483-510. Pages 488-489: Original description of male, larva, egg, pupa, and gall. (4) RiJBSAAMEN, E. H. 1892. Die Gallmiicken, des konigUchen Museums fiir Naturkunde zu Berlin. In Berl. Ent. Zeitschr., v. 37, p. 319-412. Page 375: Records synonymy and gives list of 19 species. Among them appears (as No. 7) Cecidomyia hi/pogaca Fr. Low, as mentioned by him in reference (3). (5) KlEFFER, J. -J. 1897. Synopse des Cecidomyies d'Europe et d'Algerie. In Bull. Soc. d'Hist. Nat. de Metz, Cah. 20, 2nd ser., v. 8, p. 1-64. Page 21: Mentions occurrence of subterranean galls on C. atratum and C. Icucan- themum. (6) 1897. Rectifications synonymiques. In Bui. Soc. Ent. Fr., 1897, no. 16, p. 261, par. 6. (7) 1900. Monographie des Cecidomyides d'Europe et d'Algerie. In Ann. Soc. Ent. Fr., V. 69, p. 180-472. Pages 351 and 353: Mentions injury to chrysanthemums caused hjRhopalomyia hypo- gaea F. Low. 24 BULLETIN SnZ, U. S. DEPARTMENT OF AGRICULTURE. (8) Baldrati, J. 1900. Appimti di Ceciologia. In Nuovo Giom. Bot. Ital., v. 7, n. s., p. 5-95. (No. 8G. Rhopalomyia hypogaea (F. Low) Kieff.) Pages 40-41: Describes formation and appearance of gall. (9) Kertesz, C. 1902. Catalogus Dipteronim, v. 2. Mus. Ndt. Hung., Leipsic. Page 69: Refers to changes in sjTionymy, liabitat, and hosts. (10) Lemee, E.' 1902. Alenoon. Bui. Sor. ITort., sep., p. 38, no. 131. (11) HOUARD, C. 1909. Les Zoocecidies des plantes d'Eiirope et du bassin de la Mediterran^e. V. 2, p. 988-990; v. 3, p. 1483. Fig. 1533. Describes in detail the injury caused by R. hypogaea F. Low on C. Icacanthemum, C. carymhosum, C. atratum, C. japonicum, and C. myconis. (12) KiJSTER, E. 1911. Die Gallen der Pflanzen. Pages 77 and 274: Reference to galls on subterranean roots. (13) KlEFFER, J. -J. 1913. Genera Insectorum. Diptera. Fasc. 152. Page 44-46: Refers to synonymy and describes characteristics of the genus Misospatha. (14) Felt, E. P. 1915. A new chrysanthemum pest. In Am. Florist, Apr. 10, v. 44, p. 612. Popular article mentioning its first appearance in the United States, and describing nature of injury caused by this insect. (15) (16) 1915. A new chrysanthemum pest. In Tree Talk, v. 2, no. 4, p. 27. Identical with (14) 1915. A new pest, the chrysanthemum midge (Rhopalomyia hypogaea H. Lw.). In Jour. Econ. Eut., v. 8, no. 2, p. 267. Four hosts mentioned, otherwise practically identical with information contained in (14) and (15). (17) Daniltchenko, J. N.' 1916. Chrysanthemums and their cultivation. In Garden Lib. Suppl. to Prog. Hort. amd Market Gardening. Petrograd. Resume taken from Rev. App. Ent., v. 4, ser. A, p. 164. References made to injury to chrysanthemums by Myswpatha ( Cccidomyia) hypogaea. (18) EssiG, E. 0. 1916. The chrysanthemum gall-fly, Diarthronomyin hypogaea (F. Ivow). In Jour. Econ. Ent., v. 9, no. 5, p. 461-468. Comprehensive article giving description, life history, nature of work, distribution, food plants, control, parasites and bibliography. Figs. (19) Felt, E. P. 191G. New western gall midges. In Jour. N. Y. Ent. Soc, v. 24, no. 3, p. • 175-196. Pages 192-193: Gives key to five species of Diarthronomjia, including D. hypogaea P. Low. (20) 1916. Chrysanthemum midge. In 31st Rept. State Ent. of N. Y. on injurious and other insects of the State of New York, 1915. State Mus. Bui. No. 186, p. 51-55. Fig. on p. 198. Comprehensive account giving injuries, food plants, recognition characters, technical description, life history, distribution and future probabilities, control measures, and bibliography. > Not available for reference. CHRYSANTHEMUM MIDGE. 25 (21) Gibson, A. 1916. Reports on insects of the year. Division No. 1 , Ottawa district. M 46th Ann. Kept. Ent. Soc. Ont., 1915, p. 11-14. Page 14: Notes on occurrence and control. (22) Treherne, R. C. 1916. Insects affecting agriculturists in British Columbia during the past year. In Agr. Jour. Victoria, B. C, v. 1, no. 10, Dec, p. 108. Mentions first occurrence and control. (23) Weiss, H. B. 1916. The more important greenhouse insects. N. J. Agr. Exp. Sta. Bui. 296. 42 p. Page 15: Gives r6sum6 of the salient points in Dr. Felt's article contained in Jour Econ. Ent., v. 8, p. 267. (24) Borden, A. D. 1917. Chrysanthemum ndidge. In Am. Florist, v. 48, no. 1513, p. 1061-1062. June 2. Comprehensive popular article on distribution, economic importance, life history, and control, with illustrations. (25) Gibson, A. 1917. The chrysanthemum midge. In 47th Ann. Rept. Ent. Soc. Ont., 1916, p. 118-120, pi. Comprehensive accoimt containing life historj^, habits, nature of injury, and means of control. (26) Hewitt, C. G. 1917. Report of the Dominion Entomologist for 1916. Can. Dept. of Agr. 73 p. Report of the Dominion Entomologist for 1917. Can. Dept. of Agr. 24 p. D. hypogaea contained in list of insects occurring in Canada. (27) Smith, E. D. 1917. Chrysanthemum midge or gall fly. In Am. Florist, v. 48, no. 1515, June 16, p. 1155. Illus., p. 11.54. (28) Snodgrass, R. E. 1917. Some of the important insect pests of Indiana. In 9th Ann. Rept. State Ent. Ind., 191-5-16, p. 105-230. Illus. Pages 149-151: Comprehensive article on injury, life history, distribution, and control of the chrysanthemum gall-fly. (29) Bourne, A. I. 1918. Massachusetts. In United States Dept. Agr. Emergency Ent. Service No. 11. May 1, p. 31. Mentions occurrence in Massachusetts. (30) Gossard, H. a. 1918. Ohio. In United States Dept. Agr. Emergency Ent. Service, No. 11. May 1, p. 36-37. Reports fairly satisfactory method of control by spraying with 40 per cent nicotine sulphate diluted (1-500) and fish-oil soap. (31) Guyton, T. L. 1919. Nicotine sulphate solution as a control for the chrysanthemum gall midge. Jour. Econ. Ent., v. 12, no. 2, p. 162-164. Illus. (32) Britton, W. E. 1919. Miscellaneous insect notes. In Bui. 211, 18th Rept. of State Ent. for 1918. Conn. Agr. Exp. Sta., p. 342-350, pi. 36, b. Pages 345-346: Briefly mentions occurrence with notes on life history and control. WASHINGTON : GOVEIl.VMEXT PRI.NTING OB'FICE : 1920 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 BULLETIN No. 834 Contribution from the Bureau of Entomology L. O. HOWARD. Cliief Washington, D. C. PROFESSIONAL PAPER May 19, 1920 BLACK GRAIN-STEM SAWFLY OF EUROPE^ IN THE UNITED STATES. By A. B, Gahan, Entomological Assistant, Cereal and Forage Insect Investi- gations, CONTENTS. Page. Introduction 1 History of discovery of the insect in America 2 Common name of the species 3 Distribution 3 Old-World distribution 3 Distribution in America 4 Probable future distribution in America 5 Food plants 5 Food plants in Europe 5 Food plants in America 0 Page. Synonymy and description of Tra- chelua tabidus 7 Adult 8 Overwintering larva g Manner of work and probable life history 9 Extent and character of injury n Parasites 12 Suggestions for control 12 Literature cited 14 INTRODUCTION. That another exotic insect with possibilities for damage to agricul- ture has become established in America, that it is now quite widely distributed, and that it has already begun to make itself felt was brought to light during the summer of 1918. In view of its probable future importance and the consequent desirability of learning as soon as possible the facts about its distribution, food plants, and injuri- ousness, it is deemed expedient to bring the matter to the notice of entomologists and others through publication at this time of such information as is at hand.^ ^Trachelus tabidus (Fab.). 2 The writer is Indebted to Miss Margaret Fagan, of the Bureau of Entomology, for compilation of the appended bibliography ; to Messrs. W. R. McConnell, P. R. Myers, and W. J. Phillips, also of the Bureau of Entomology, for reared and collected material and notes which were invaluable in establishing the identity and distribution of the insect ; to Dr. J. Chester Bradley, of Cornell University, for larval material of Cephtis pygmaeus as well as useful suggestions regarding characters for separating the larvae ; and to Messrs. S. A. Rohwcr and Wm. Middleton for numerous helpful suggestions. 148465°— 20— Bull. 834 1 2 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. HISTORY OF DISCOVERY OF THE INSECT IN AMERICA Specimens of a sawfly were collected by C. W. Johnson at River- ton, N. J., some time prior to 1899, tlie exact date unknown. These were submitted to Dr. W. H. Ashmead for determination and pro- nounced by him a new species, to which he gave the manuscript name of Calajneuta johnsonL Under this name, misspelled Calamen ta john- soni, the record was published, without description, in the second edi- tion of Insects of New Jersey, by John B. Smith (39) .^ A description based on these specimens was later published by Ashmead (40) . Sub- sequently the type specimens were examined by Dr. J. Chester Brad- ley, who recognized in them the European species Trachelus tahkJus Fabricius, and under this corrected name the record of the original collection was again published in the third edition of Insects of New Jersey (45). At that time, and for some time thereafter, nothing was known of the food i)lants of the species in this country. Such European records as existed were in the Russian literature, and be- cause of difficulty in translation were largely overlooked. Since Trachelus tahidus was not known to be causing any injury, no par- ticular attention was paid to it. During the summer of 1918 a complaint was received by the Bu- reau of Entomology from a correspondent at Gaithersburg, Md., re- garding the work of some insect which had caused his ripening wheat to fall badly, and the writer was detailed by Mr. W. R. Walton, in charge of Cereal and Forage Crop Insect Investigations, to investi- gate the outbreak. Specimens of the injured stalks were received from the corre- spondent, some of which proved to contain larvae of what was readily determined as a species of Cephidae. The insect was at first thought to be either CeyKus pygmaeus Linnaeus or Cephus cinctios Norton. The former species had been known in the vicinity of Ithaca, N. Y., many years before, but so far as the records indicated, had succeeded in spreading but little from the point of original infestation and had never been recorded as seriously injurious. Cephus clnctus was known only from the western United States and was not believed to occur east of the Mississippi River except for a few localities in Michigan. Wlien the Maryland infestation was brought to the attention of Mr. S. A. Rohwer, specialist on sawflies, of the United States Depart- ment of Agriculture, Bureau of Entomology, he at once suggested the possibility that the insect might be Trachelus tahidus Fabricius. Several adult specimens of this species in addition to those originally taken at Riverton, N. J., were in the United States National Museum. These, as shown by the labeling, were collected at two or three differ- '^ Numbers in parenthesis refer to " Literature cited," p. 14. BLACK GRAIN-STEM SAWFLY. 3 ent points in Maryland and Pennsylvania. The fact that Trachelus tabidus was thus known to occur in the region of the infestation and that it was the only species of Cephidae known to occur there which would be likely to have the habit attributed to this one, at once es- tablished a strong probability that it was the species concerned. This suspicion was confirmed when Messrs. W. K. McConnell and P. R. Myers, who had been engaged for several years in investigating the insect enemies of wheat and other cereals for the Bureau of Ento- mology in Maryland and Pennsylvania, turned over to the writer all of the sawfl}^ material which they had secured from that region. This material all proved to be Trachelus tdbldus and included sev- eral specimens which had actually been reared from wheat stubble in breeding cages, as well as numerous specimens collected by sweep- ing wheat. The reared material left little doubt as to the identity of the depredator in question. Reference to the European literature, especially that by a number of the more recent Russian entomologists, regarding Trachelus tahldus^ cleared up whatever lingering doubts might still have existed. The manner in which this insect became established in America is unknown and probably will remain a mystery. Any surmise as to the probable manner of introduction would be valueless and since it could have no effect upon the fact that it probably has come to remain, is omitted. COMMON NAME OF THE SPECIES. In Russian literature this species is referred to as the black sawyer or black sawfly because of its nearly uniform black color. The writer has ventured to modify this name to some extent by calling it the black grain-stem sawfly. This name is descriptive of the insect as well as of its habit and at the same time minimizes the possibility of confusion with other black species of sawfly. DISTRIBUTION. OLD-WORLD DISTRIBUTION. Andre (33) records the species as occurring in England, France, Sweden, Spain, Germany, Italy, Algeria, and Syria. Cameron (35) in his " Monograph of the British Phytophagous Hymenoptera " gives practically the same distribution, omitting France. Konow (42) in Genera Insectorum cites middle and southern Europe, Algeria,, and Asia Minor. Several writers, including Shtchegolev (49) , Kurd- jumov (48), Uvarov (50, 54)', Zolotarevsky (53), and Borodin (52), have recorded it more recently from central and southern Russia, while Kulagin (47) states that Cefhus pygmaeus and Trachelus tabidus have been recorded from 26 different governments of Russia. 4 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. From this it Avill be seen that the species occurs in all three Old World continents, being present throughout most of the middle and southern European countries, southeastern Asia, and northern Africa. It is probable that it will eventually be found to be present through- out most of the remaining re'gions where wheat is gi'own in Europe and eastern Asia, as well as northern Africa. DISTRIBUTION IN AMERICA. The present distribution in America, gleaned from the writer's in- vestigations, the records of Messrs. McConnell and Myers, which were kindly placed at his disposal, specimens in the United States National Museum, and a note furnished by Dr. J. Chester Bradley, includes localities in six States, viz : New Jersey, Virginia, Delaware, Maryland, Pennsylvania, and New York. The localities where the insect is at present known to occur, to- gether with the date of collection and the name of collector, are as follows : Riverton, N. J., adults collected by C. W. Johnson, May 29 (year not given but prior to 1899) ; East Falls Church, Va., hibernating larvae in wheat and rye stubble, A. B. Gahan and S. A. Rohwer, collectors, August 6, 1918; War- renton, Va., larvae collected in August, 1918, by W. J. Phillips; Harrington, Del., larvse collected in wheat stubble by W. R. McConnell; Wolfsville, Md., adults swept in clover field by J. A. Hyslop, June 6, 1914; Hagerstown, Md., adults swept from wheat by P. R. Myers and W. R. McConnell, May 15, May 25, June 4, 1915, and May 24 and June 8, 1916; Hagerstown, Md., adults reared from wheat stubble by P. R. Myers, May 10, 1916, and by W. R. McConnell, May 3, 1917; Taylors Island, Md., adult reared from wheat stubble by W. R. McConnell, March 21, 1917 ; Great Falls, Md., adult swept by Fredericli Knab, May 24, 1914; Germantown, Md., larvse injuring wheat received from a cor- respondent by the Maryland State College of Agriculture, and recorded under the name of Ccphiis pygmacus in Reiwrt of Maryland State Horticultural Society, 1914 ; Gaithersburg and Laytonsville, Md., hibernating larvfe collected by A. B. Gahan in wheat stubble, July 22, 1918 ; College Park, Md., hibernating larvse in wheat by A. B. Gahan, July 24, 1918; Berwyn, Md., hibernating larvse in rye stubble collected by A. B. Gahan, July 30, 1918; Laurel, ]\Id., hibernating larvse in wheat stubble, by A. B. Gahan, July 26, 1918; Towson, Md., hibernating larvse in wheat stubble, by A. B. Gahan, August 2, 1918; Warfordsburg, Pa., adult reared from wheat stubble by P. R. Myers, June 5, 1916; Hunters Run, Pa., adult reared from wheat stubble, W. R. McConnell, May 7, 1918; West Chestei*, Pa., adult collected by H. L. Parker, .June, 1915; Linglestown, Pa., adults collected by H. B. Kirk and A. B. Chaniplain, May 26, 1908 ; Herndon, Pa., adults, collector unknown, June 6, 1907 ; State College, Pa., adult swept in clover by W. R. McConnell, May 30, 1911 ; Carlisle, Pa., larvse collected July 23, 1918, by C. C. HiU, and July 31, 1918, by W. R. McConnell ; Mt. Holly Springs, Pa., larvse collected in wheat stubble, Juno 25, 1918, by McConnell and Myers. In addition to these records Dr. J. C. Bradley states that he has seen one adult specimen in the collection of the Brooklyn Museum fi'om Long Island, N. Y. :*l BLACK GKAIN-STEM SAWFLY. These records probably indicate only approximately the present distribution (fig. 1) of the species. A more extended investigation will be necessary to establish the exact limits. The records are sufficient to establish a probability that it already occurs over the greater part of Virginia, Maryland, Pennsylvania, New Jersey, and Delaware, and it is possible that AVest Virginia, eastern Ohio, southern New York, and even some of the New England States may already be included within its range. PROBABLE FUTURE DISTRIBUTION IN AMERICA. Judging by its wide distribution in the Old World, as well as by the character of localities already included within its range in this country, including both tidewater and mountain districts, there seems little reason to doubt that the species will eventually spread over all the wheat- growing sections of the eastern and central United States. Whether it will accommodate itself to the arid and semiarid wheat- growing districts of the West is matter for specu- lation. The average pre- cipitation of this region does not differ greatly from that of southern Russia, where the species seems to be at its worst. The mean temperatures of the two regions probably are not widely differ- ent. It seems possible, therefore, that unless some other climatic or physical factor intervenes, the species may spread eventually from coast to coast. That it will spread northward into Canada may be doubted, since in Europe, although recorded from Sweden, it seems not to occur generally in the colder northern portion. FOOD PLANTS. FOOD PLANTS IN EUROPE. In European literature, with one exception, the records of food plants of Trachelus tahidus are by Russian entomologists. Rudow (43), in speaking of Cephus satyi^s Panzer, C. nigrinus Thomson, C. pallipes Klug, C. arundinis Giraud, and C. tahidus Fabricius, makes the general statement that in southern Europe these species live in Fig. 1. -Present known distribution of Trachelus tahidus in America. 6 BULLETIN 834, IT. S. DEPARTMENT OF AGRICULTURE. Triticum repens^ Bromus, Holcus, and reeds or bulrushes, without specifying which plants serve as host plants for the different species. The record is, therefore, too general to be of much value in the present circumstances. The Kussian records are more specific and apparently all refer to infestation of small-grain crops. Shtchegolev (49) records the species as a serious pest of wheat; Borodin (52) reared it together with Cephus 'pygmaeus from barley and both spring sown and winter sown wheat; Kulagin (47) states that it infests rye and wheat in the Province of Taurida. Various other Eussian writers, includ- ing Uvarov (50) (54),Zolotarevslr7 (53), and Kurdjumov (48), have recorded the species as attacking the grain crops, principally wheat. FOOD PLANTS IN AMERICA. Thus far in America only wheat and rye are known to serve as food plants. Only two instances of its occurrence in rye are known, both observations having been made by the writer during the past summer. In each instance in which it was found in rye only one infested stalk was located in the field, although careful search was made. Although several fields of oats stubble were examined more or less carefully, no infestation of this crop was observed even when the oats were in close proximity to infested wheat fields. No oppor- tunity to examine barley has offered itself. It appears fairly certain that of the cultivated grain crops wheat will prove to be its preferred food plant. As stated elsewhere, the writer was able to find larv89 in varying quantities in every wheat field visited in the course of the limited investigation carried on during the summer of 1918. That the species eventually will be found to infest plants other than the small grains is possible. The related species, Cephus cinctus, which also infests the small-grain crops in our Western States, is known to have a long list of food plants among the coarser grasses, both wild and cultivated. Cephus cinctios is believed, how- ever, to be a native species whose original food plants were wild grasses and whose habit of attacking the cultivated grains has been induced through elimination or reduction of its natural source of food supply by the development of agriculture. Present informa- tion indicates that the natural food plants of Traclielus fahidus are the small grains, although it may yet prove to have other hosts in Europe. It is doubtful whether the species which has been trans- planted from its native habitat to America will find it possible to develop in our native grasses. Only future investigation can de- termine. The possibility of its being able to maintain itself upon grasses other than the small grains should be borne in mind, and BLACK GRAIlSr-STEM SAWFLY. 7 such grasses as Elymus, Bromus, Agropyron, and in fact any of the cultivated or wild grasses whose seasonal development is such as to provide a succulent stem of sufficient size to accommodate the larvse during the period of larval activity, May and June, should be viewed with suspicion. SYNONYMY AND DESCRIPTION OF TRACHELUS TABIDUS. Since the original description by Fabricius (1), the name has undergone a number of generic transfers, and it has been redescribed or referred to under several specific names. It is the opinion of the writer, however, that some of the pub- lished synonymy is incorrect. TracTielus JiaemoTroidalis Jurine (17) is listed as a synonym by Dalla Torre (38). The figure given by Jurine is not that of tahldus^ but is undoubtedly identical with Cephus analis Klug (12). The latter name is generally recognized by European writers as a synonym of haeinorrholdalis Fabricius (2). Konow (41) includes Sirex macilentus Fabricius (9), Cephus erheri Damianitsch (30), and Cephus vittatus Costa (32) as color varie- ties of tdbidus. The large number of specimens of tabldus examined by the writer show no variation in color comparable to m/wilentus as illustrated by Coquebert (10), and since other European writers generally have recognized this as a distinct species it has been ex- cluded from the synonymy. Cephus erberl is described as having the third to seventh dorsal abdominal segments banded with yellow, the band on the third segment interrupted, while the third to ninth ventral segments are said to be margined with yellow laterally. In no specimen of tdbidus examined by the writer is there any indica- tion of a dorsal band of yellow^ on the abdomen nor are the ventral segments margined with yellow. Like macilentus^ this species is therefore excluded from the synonj^my. The description of vittatus^ on the other hand, agrees closely with tabldus^ and there seems to be no good reason why it should not be considered a pure synonym in- stead of a color variety. Following is the complete synonymy of TracJielus tabldus as understood by the writer: Traclielus tajbidus (Fabricius) Jurine. Hirex tabidus Fabricius (1), (2), (4). — Villers (5).— Linne (6).— Christ (8). — Fabricius (9). — Coquebert (10). — Walckenaer (11). — Rossi (19).— Herricli-SchafEer (26).— Dumeril (29). Tenthredo longicoUis Fourcroy (3). — Villers (5). Astatus tabldus Klug (12). — Panzer (15). Cephus tabldus Fabricius (13). — Latreille (14). — Panzer (16). — Latreille (18).— Lepeletier (20).— Stephens (23).— Hartig (24).— Blanchard (25).— Lucas (27).— Costa ( 28 ) .— Taschenberg (31).— Andrg (33).— Cameron (34), (35).— Magretti (36).— Costa (37).— Dalla Torre (38). 8 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE, Trachelus tabidus Jurine (17). — Curtis (22). — Kouow (41), (42). — Rudow (43).— Kokujev (44).— MacGillivray (45).— Kliolodkovsky (4G).— Kulagin (47). — Kurdjumov (48). — Shtcliegolev (49). — Uvarov (50).— Borodin (52). — Zolotarevsky (53). — Uvarov aud Glazuuov (54). — Mac- Gillivray (55).— Howard (56). CepJius nigntus Lepeletier (20). Ceplius vittatus Costa (32), Calamenta johnsoni Aslimead (39). Calamcnta johnsoni Aslimead (40). Ceplius pyginaeus Cory (51), not Linnaeus. ADULT. Female. — Long and slender, with abdomen somewhat compressed. Antennae longer than thorax, thickened at apex, thii'd joint slightly shorter than fourth, third to eighth joints four to five times as long as thick, those beyond eighth gradually shortening, the ones in thickened portion of antennte subequal and not longer than broad ; clypeus truncate at apex, with sharp lateral angles ; occiput concave ; greatest width of posterior orbit about equal to greatest trans- verse diameter of eye ; pronotum elongate, nearly as long as broad, broadest at posterior margin, its sides more or less concave; mesoscutum and scutellura subequal in length ; middle and hind tibise each with t\vo apical spurs and normally each with two (sometimes one only) superapical spurs located at about the apical one-third of tibiie ; abdomen much longer than head and thoi'ax, subcompressed, with sheaths broad and slightly thickened at apex. Polished black ; mandibles except at apex, small spot below tegulse, apex of front femora within, inner side of front tibiae for its whole length, spot at apex of median femora in front, a broad longitudinal stripe along lateral margins of dorsal segments, a lateral spot on seventh ventral segment, and a narrow marginal stripe on basal part of sheaths yellow; wings subhyaline, nervures and stigma black. (See PI. II, A.) Male. — Similar to female in general appearance and color. Last visible ventral segment of abdomen prominent and extending beyond last dorsal seg- ment ; two ventral segments before last visible segment each with a broad, deep, horseshoe-shaped depression, within which is a transverse row of stiff, erect bristles. (See PI. II, B.) The only species occurring in America, so far as known, with which this species is likely to be confused are CepJviis cinctus Norton and Cephus pyginaeus Linnaeus. The female may be readily dis- tinguished by the fact that in both cinctus and pygmaeus the sheaths are narrower at apex than in the middle, and the dorsal segments are banded apically with yellow. The legs are also more largely yellowish. The male of tahidus differs from the males of the other two species in the presence of the horseshoe-shaped depression on the two ventral segments as well as by the color characters pointed out for the female. OVERWINTERING LARVA. Body subcylindrical, yellowish white in color, 7 to 9 mm. in length, thoracic region slightly thicker than abdominal region. Living larvae more or less tinged with greenish due to body contents. Head pale yellow, its dorso-ventral length 1.1 mm., breadth 1.1 mm. Anten- nae 5-jointed, tapering to a point at apex, more or less fuscous in color. Man- dibles four-toothed, brownish, their apices nearly black; labrum, sutures about Bui. 834, U. S. Dept. of Agriculture. Plate I. The black Grain-Stem Sawfly of Europe in the United States. A, Trachelus iabidm: Lateral view of apical segments of larva; B, Cephus cinctns: tate-al view of apical segments of larva; C, Cephus pygmaeus: Lateral view of apical segments of larva. Bui. 834, U. S. Dept. of Agriculture. Plate 1 1 . Trachelus tabidus. A, Adult female: a, Lateral view of female abdomen. B, Adult male: 6, Ventral view of male abdomen. BLACK GRAIN-STEM SAWFLY. 9 clypeus, and articulations of mandibles brownish. Thoracic legs small, papilli- form, and pale brownish. Dorsal abdominal segments triannulate; pleura prominent; ventral segments triannulate; prolegs absent. (For further de- scription see Plate I, A, and key to larvae.) Larvse of this species are apt to be confused with those of Cephus pygmaeus and C. cinctus. All three species infest the small grain crops in practically the same manner, have very similar biologies, and superficially resemble one another closely. The following key in conjunction with the accompanying figures will, it is believed, make it possible to recognize them : KEY FOR SEPARATING GBAIN-INFESTING SAWFLY LAEV^. 1. Dorsal anal lobe of the tenth tergite viewed from the side triangular, slop- ing gradually from base to apex, the anterior end of lobe much thicker than the posterior end which is more or less acute. Spines on the anal prong each arising from a small, more or less chitinized tubercule and closely grouped about the apex of the enlarged fleshy portion just basad of the short chitinized apical ring. Eighth and ninth tergites ap- parently' glabrous 2. Dorsal and lobe viewed from the side not triangular, not sloping gradu- ally from base to apex but convexly rounded, the posterior end of the lobe as thick dorso-ventrally as the anterior end and nearly perpendicu- lar. Anal prong completely encircled by two irregular series of widely separated, short, stiff spines which do not arise from chitinized tubercules. Eighth and ninth tergites each with a transverse row of distinct short hairs. (See Plate I, A.) Trachelus tabidvs Fab. 2. Anal prong terminating in a short chitinized ring which is not as long as broad. Spines basad of chitinized ring few in number, confined to a single transverse row on the dorsal surface. Dorsal, lateral, and ventral lobes all sparsely hairy. (See PI. I, A.) Cephus pygmaeus Linn. Anal prong terminating in a chitinized tube-like process which is distinctly longer than broad. Spines basad of the apical tube-like process numerous, arranged in two irregular contiguous series completely encircling base of tube. Anal lobes all more distinctly hairy. (See PI. I, A.) Cephus cinctus Nort. MANNER OF WORK AND PROBABLE LIFE HISTORY. Observations by the writer supplemented by those of McConnell and Myers, and confirmed in part by Kulagin (47) indicate that the life history of Trachelus tabidus does not differ greatly from that of the western grass-stem sawfly, Cephus cinctus. The early stages have not yet been observed in this country. Collections of adults in the field indicate that egg-laying takes place during the period May 15 to June 10. Some of the emergence records for reared specimens show somewhat earlier dates than May 15, but since these were obtained under the abnormal conditions of the labora- tory they may be disregarded. Kulagin indicates about the same period for egg-laying in Russia; 148465°— 20— Bull. 834 2 10 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. The eggs are inserted in a slit made by the female in the stem some distance above the ground. The young larvae burrow down- ward through the pith of the stem, hollowing it out to the base. The larvae apparently attain full growth at about the time the grain is ripe. At the time the writer's investigations were begun, July 22, they were full grown and had evidently gone into hiber- nation. The fully-developed larvae were, at this time, located at the extreme base of the stem, encased in a silken tube or lining to the burrow, this tube being two or three times the length of the larva and filling the hollow straw completely from about the surface of the ground downward. Above this silken tube the bur- row is completely closed by a wad of frass. Before this wad of frass is put in place the larva almost completely severs the stem from the inside, this cut in nearly every c^se being at or very near the surface of the ground and usually a little above the first node on the stem where the surface roots put out. In making this cut just enough of the epidermis of the straw is left unsevered to attach it lightly and allow it to stand erect. In consequence of this cut the first slight bending of the ripened straw, as by a strong wind, causes it to snap off and fall. Straws cut off by the insect are easily distinguishable from those severed by the harvester. The cut is exactly transverse to the stem and the ends of both the stub which remains in the ground and the fallen straw are distinctly concave or funnel-shaped. This appear- ance is so characteristic that one can readily detect the presence of the pest in a field by simply examining the ends of the fallen straws. The hibernating larva, as already stated, is to be found in the stub of the wheat stalk remaining in the ground. Since this stub in most cases barely extends to the surface of the soil (sometimes a little above the surface) , it is not always an easy matter to locate it. The fallen straws sometimes remain slightly attached on one side to the stub, and in such instances one can locate the infested stub by following the straw to its base. When the straw is completely de- tached, as is usuall}' the case in a field that has been harvested, it is often necessary to search for some time before the stub containing the larva can be located. The greater part of the insect's life cycle is apparently spent as a hibernating larva, this period extending from about the time the wheat is ripe enough to cut until some time the following spring when the larva changes to the pupa. Just when this change takes place has not been ascertained, but the pupal state is probably of short duration, as in Cephus cinci/m; if so, the change to the pupa probably occurs in the latter part of April or early part of May. Adults, as already stated, are to be found in the fields during the latter half of May and early June. BLACK GRAIN-STEM SAWFLY. 11 In the case of one male specimen reared from wheat stubble by McConnell, the insect was collected as a larva September 30, 1915, and emerged as an adult May 3, 1917. This shows that under some conditions the life cycle may be extended over a two-year period. Such instances are probably rare under natural conditions, but the record indicates a high degree of adaptability on the part of the species for overcoming unfavorable conditions. EXTENT AND CHARACTER OF INJURY. Shortly after receipt of the complaint from Gaithersburg, and be- fore the identity of the insect was known, the writer was detailed to visit the locality and investigate the nature and extent of the damage. Accordingly, on July 22, 1918, the farm of the correspondent, Mr. Beverly R. Codwise, was visited. It lies about 3 miles north of Gaithersburg on the road to Laytonsville, Md. Unfortunately the infested fields had already been harvested, making it practically im- possible to estimate the extent of the actual injury. It was evident, however, that there had been an appreciable loss due to falling of the grain, so that it could not be picked up by the binder. The writer was informed that comment by passers-by on the large amount of grain missed by the binder had first called attention to the injury and caused the investigation which resulted in the sending of samples to the Bureau of Entomology. Three other farms in the neighborhood of Gaithersburg were visited, on each of which the pest was located. Subsequently wheat fields in various other localities in Maryland and one in Virginia were visited, with the result that in every wheat field examined the insect was found to be present in varying abundance. No actual counts of infested straw were made by the writer in any of the fields to ascertain the percentage of infestation. It was roughly estimated that in some of the worst cases the infestation amounted to 4 or 5 per cent. In most of the fields the infestation was much less than 4 per cent. Messrs. McConnell and Myers did make counts of the infested stubble on a small number of experimental plats and other fields of wheat at Carlisle and Mount Holly Springs, Pa., with the results showing an infestation varying from 4.36 per cent on one of the plats to 0.26 per cent on another, the aveo^age from all counts being 1.75 per cent. That a 4 per cent infestation of stubble necessarily indicates a 4 per cent loss in the crop is not probable. The insect apparently chooses only well-developed and vigorous stems in which to oviposit, and some of these infested stems are known to develop at least par- tially filled heads. Just what the loss, if any, from failure of the heads on infested stalks to fill properly may be, remains to be de- termined. 12 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. It is certain that serious loss may and does occur because of fall- ing of the grain due to the cut made by the larva preparatory to hibernation. The extent of this loss will in all probability depend in some degree upon weather conditions during the period of ripening of the grain. A heavy wind or severe storm at this time would cause most of the infested grain stalks to break off and fall so that they would not be picked up by the harvester. In the absence of such a wind or storm, the loss would undoubtedly be much less, but even with favorable conditions a certain percentage of the infested stalks would be broken off by the harvester reel and fall in front of the plat- form. Kulagin (47) states that losses in Kussia due to this species and Cefhus fygmaeus are estimated at 14 to 20 per cent, although more severe in some cases. Shtchegolev (49) also reports 15 to 20 per cent injury due to these two pests. Other Russian writers record severe injury without specifying the amount. Unfortunately, in practically every instance these writers treat of the injury by CefJms fygmaeus and Trachelus tahidus collectively, without indi- cating how much of the damage is chargeable to each. This is no doubt due to the fact that they have been unable to distinguish the larvse of the two species. Their records, therefore, do not afford a reliable basis for estimating the probable future importance of Trachelus tahidus in this country. PARASITES. In Russia two parasites of Traehelws tdbidus have been recorded. CoUyria calcitrator (Gravenhorst), an ichneumonid wasp, is ap- parently a common parasite of this species as well as of Cephus fygmaeus. Borodin (52) records the chalcidid Arthrolysis {Picro- cystus) scabricula Nees as having been reared from these two saw- fly pests. Neither of these species has, as yet, been found in America. The existence of at least one efficient parasite in America has, however, been established. Numerous specimens of a chalcidoid belonging to the genus Pleurotropis and apparently representing an undescribed species have been reared by Mr. W. R. McConnell at Mount Holly Springs and Carlisle, Pa. So far little is known of the life history of the species. It emerges from the prepupal larva of the Trachelus at about the time of emergence of the host adults and is believed to be a primary parasite, solitary in its habit. Observ ations to date are too limited in extent to form a very accurate estimate of the efficiency of this parasite, but in some instances at least it appears to exercise a considerable degree of control. SUGGESTIONS FOR CONTROL. In the present state of our knowledge of this species only sug- gestions of possible means of control can be given. BLACK GRAIN-STEM SAWFLY. 13 Measures for control will doubtless be similar to those against Cephus cinctm, the western grass-stem, sawfly, and Cephus pygmaeus. It is obviously impracticable to attack the insect in the egg stage or active larva stage, since both stages occur in the growing grain. The adult can not be reached by any known method. It appears, therefore, that control measures to be successful must either aim at destruction of the hibernating larva or pupa while in the stubble, where the insect passes the greater part of its existence, or be con- fined to cultural methods such as crop rotation. The fact that the larva is located in the part of the stem below^ the surface of the ground precludes the possibility of accomplishing anything by any ordinary burning of the stubble. Disking the stubble thoroughly as soon after cutting the grain as practicable would possibly be of benefit by turning the infested stubble out and thus exposing the larva to the action of summer heat and winter cold. The larva is quite hardy, however, and only experimentation will prove or dis- prove the effectiveness of disking. In Russia plowing down of the stubble as deeply as possible is recommended against this species as well as against Cephus pyg- maeus. The same treatment is recommended and has proved success- ful against Cephus cinctus in this country, and it seems the logical treatment for adoption against this species. To be effective the plow- ing must bury the stub containing the larva so deeply that the matur- ing adult will be unable to escape. Shallow plowing will not suffice, as the adults undoubtedly will be able to burrow their way out if covered with only 2 or 3 inches of soil. Deep and clean plowing, therefore, will be essential. The plowing may be done at any time between the cutting of the grain and the following spring prior to emergence of the adults in April or May. If, as now seems probable, this insect confines itself, in this coun- try, to the small-grain crops as host plants, there can be little doubt that crop rotation w^ill prove an effective means of reducing damage from it. Wheat, barley, or rye should be followed by some crop which will not serve as a host plant, such as corn or truck crops. The present practice on some farms of growing wheat on the same ground two years in succession is distinctly favorable to the propagation of the pest since the adults upon emerging find themselves surrounded by ideal conditions for oviposition. The practice also of sowing grass or clover with wheat and allowing the wheat stubble contain- ing the larva to stand undisturbed for two seasons could hardly be improved upon as means of increasing the numbers of this insect. Any system of rotation to be effective should insure thorough plow- ing down of the wheat stubble and the growing of some crop other than a small grain following wheat. 14 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. LITERATURE CITED. (1) FABRicrus, J. C. 1775. SYSTEMA ENTOMOLOGiAE . . . Flensburgi et Lipsiae. S32 p. Page 326, no. 8 : Original description under name of Sirex taiiduK, England. (2) 1781. SPECIES INSECTORUM . . . Hamburgi et Kilonii. Tome 1. Page 420, no. 13 : Brief description of Sirex tabldus, with note, " Habi- tat in Angliae floribus." Page 417, no. 59 : Original description of Tenthredo h-aemorrliotdalis. (3) FOTJBCEOY, A. F. DE. 1785. ENTOMOLOGiA PARisiENSis . . . Patis. Fars 1. Page 378 : Describes species as new under name of Tenthredo longtcolli» from vicinity of Paris, France. (4) Fabkicius, J. C. 1787. MANTISSA INSECTORUM . . . Hafuiae. Tome 1. Page 259, no. 19 : Brief description of Sirex toMdus. (5) ViLLERS, C. DE. 1789. CAROLi LiNNAEi ENTOMOLOGIA . . . Lugtlunl. Tome 3. Page 125, no. 135 : Record, as TentJii-edo longicollis, from Europe. Page 132, no. 13 : Short description of Sirex tahidus, with note, " Habitat in Angliae floribus." (6) LiNNf;, C. 1790. SYSTEMA NATURAE . . . E(l, 13 (Gmelin), Lipsiae. Tome 1, pars 5. Page 2674, no. 18 : Short description of Sirex tattdus, with note, " Habi- tat in Angliae floribus." (7) Rossi, P. 1790. FAUNA ETRUSCA . . . Liburiii. Tome 2. Page 35, no. 730 : Records of Sirex taMdiis from Italy. (8) Christ, Johann Ludwig. 1791. naturgeschichte, classification und nomenclature dee insecten. . . . Frankfort-am-Main. 535 p. Tab. 60, col. Page 416 : Sirex tahidus. (9) Fabeicius, J. C. 1793. ENTOMOLOGIA SYSTEMATICA . . . Hafniae. Tome 2. Page 131 : Short description of Sirex taiiduo, from England ; original description of Sirex nidcilentus. (10) COQUEBERT, A. J. 1801. iLLUSTRATio icoNOGRAPHicA INSECTORUM . . . Paris. Tome 2. Page 48, Tab. 11, fig. 4 : Figure and brief description of Sireic taMdus^ from England. (11) Walckenaer, C. A. 1802. FAUNE PARisiENNE, iNSECTES. Paris. Tome 2. Page 46 : Record of Sirex tahidus from vicinity of Paris. (12) Klug, Fr. 1803. monographia sibicum geemaniae . . . Berlin. 83 p. Page 56, no. 8 : Generic transfer with description and synonymy of Astatus tahidus and note, " Locus in floribus." Page 55, no. 6, Tab. 7, fig. 1 : Original description of Cephus analis. (13) Fabricius, J. C. 1804. SYSTEMA piezatoeum . , . Bruiisvigiae. 439 p. Page 252, no. 6: Transferred to Cephus tahidus with references to previous literature. Locality, England. (14) Latreille, p. A. 1805. histoire naturelle . . . cbustac:6s et des insectes. Paris. Tome 13. Page 144 : Record of Cephus tahidus as common around Paris. BLACK GRAIN-STEM SAWFLY. 15 (15) Panzee, G. W. F. 1805. FAUNA iNSECTORUM GERMANicAE . . . Niimberg. Heft 85. Tab. XI : Figure of Astatus tahidiis, I'eferences to previous literature and note, " Habitat in sylvaticis." (16) 1806. KRITISCHE REVISION DEE INSEKTENFAUNE DEUTSCHLANDS . . . Niirnberg. Bd. 2. Page 145 : References to previous literature on Cephus tabidus. (17) JuRiNE, Louis. 1807. NOUVELLE METHODS DE CLASSER LES HYMIENOPTEEES ET LES DIPT^EES. Geneva. Tome 1. Page 72: Description of Trachelus lujcmorroidalis ; short description of Trachclus tahidus. (18) Lateeille, p. a. 1807. geneea crustaceoeum et insectorum . . . Paris. Tome 3. Page 236 : Listed as Cephus tabidus. (19) Rossi, P. 1807. fauna eteusca . . . Helmstadii (Ed. Illiger). Tome 2. Page 53 : Brief description of Sirex tabidus with notes, " In Gallia meridionali ..." " Hab. in floribus rarior." (20) Lepeletiee de St. Faege^au, Am. 1823. MONOGEAPHiA tenthbedinetaeum . . . Paris. 176 p. Page 19, no. 54 : Species described as new under name of Cephus mandibularis, with note, " In agro Parisiensi." Page 20, no. 55 : Species described as new under name of Cephus nigritus, with note, "An mas Cephi mandibularis f " Page 20, no. 57 : Description of Cephus tabidus with references tO' previous literature. (21) Altdinet-Seeville, J. G. In: Vieitxot, Lepeletiee, etc. 1823. faune fbanoaise . . . Paris, Hyra6nopt§res. Page 90, no. 5 : Cephus mandibularis. (22) Curtis, .John. 1830. BRITISH entomology . . . London, v. 7. Plate 301 (accompanying text) : Trachelus tabidus recorded on flowers and grass in fields of Dover and Southgate, England. (23) Stephens, J. F. 1835. ILI.USTRATIONS OF BRITISH ENTOMOLOGY . , . Mandibulata. Lon- don. V. 7. Page 106, no. 9: Description of Cephus tabidus; found in Hertford, Cambridge, Whittlesea Mere, Dover, Southgate, England. (24) Hartig, Theodor. 1837. Die Familien der Blattwespen ltnd Holzwespen . . . Berlin. 416 p. Page 363 : Description of Cephus tabidus. (25) Blanchard, C. E. 1840. histoire naturelle des insectes . . . Paris. Tome 3. Page 224, no. 3 : Description of Cephus mandibularis ; no. 4 : Descrip- tion of Cephus tabidus; found near Paris. (26) Herrich-Schaffer, G. A. W. 1840. nomenclatoe entomologicus. Regensburg. Heft 2. Pages 92-93 : Lists .synonymy of Sirex tabidus. (27) Lucas, H. 1849. histoire naturelle des animaux articul^s. Pars 3. Insectes . . . Paris. In Exploration Sci. de I'Algerie . . . 1840, 1841, 1842, Page 342 : Records Cephus tabidus taken in May in Constantine. 16 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. (28) Costa, A. 1860. FAUNA DEL REGNO Di NAPOLi . . . NapoU. Iiueuotteri. Page 8 : Cephus tahidus. (29) DuM^RiL, A. M. C. 1860. ENTOMOLOGiE ANALYTiQUE. Paris. 1339 p. Page 080 : Note that larva? of Sirex tabidus live In stems of grain. (30) Damianitsch, Rudolph. 1866. HYMENOPTEROLOGiscHE BEiTRAGE. In Verb. Zool.-Bot. Ges. Wien, Bd. 16, p. 993-996. Pages 993-995 : Original description of Crphus erieri, taken at edge of barley field in Syria. <31) Taschenberg, E. L. 1871. EINIGE NEUE SUDEUROPAISCHE HYMENOPTERA. Itl Zoitschr. f. d. Ges. Naturw., Bd. 4 (n. s. ). Page 306 : Description of Cephus tabidtis ; found in Island Lesina. (32) Costa, A. 1878. REI.AZIONE DI UN VIAGGIO PER l'EGITTO, LA PALESTINA, E LE COSTE DELLA TURCHIA ASIATICA PER RICERCHE ZOOLOGICHE. In Atti ACC. Sci. Napoli, v. 7, no. 2. Pago 14 (footnote) : Species described as new under name of Cephus littatusj from Egypt. <33) ANDRfi, Ed. 1881. SPECIES DES HYMfiNOPTJ&RES d'eUROPE ET D'ALG^RIE . . . BeaUDG, Cote-d'Or, Tome 1. Page ."5.3.5, no. 31 : Description of Cephus tahidus ; found in England, France, Spain, Switzerland, Algeria, Italy, Germany, Syria. <34) Cameron, P. 1886. A synopsis of the British species of cephina. In Ent. Mo. Mag., V. 22, p. 17.5-177. Page 177 : Brief description of Cephus tahidus j synonymy ; table. (35) 1890. A monograph of the BRITISH PHYTOPHAGOUS HYMENOPTERA. London, v. 3. Page 120, no. 6, PI. 3, fig. 9 : Description, previous references in litera- ture, and synonymy of Cephus tabidus. Found In England, Sweden, Germany, Spain, Italy, N. Africa, Syria. (36) Magretti, P. 1890. iMENOTTERi DI siKiA . . . In Ann. Mus. Civ. Stor. Nat. Genova, (2) V. 9, p. 522-548. Page 528 : Cephus tabidus recorded from Europe, Algeria, Syria. ■(37) CoSTA, A. 1894. PROSPETTO SISTEMATICO DEGLI IMENOTTERI ITALIANI. Napoll. ParS 3. Page 253 : Cephus tabidus recorded from Italy and Sicily. (38) Dalla Torre, C. G. de. 1894. catalogus hymenopterorum . . . Lipsiae. v. 1. Page 412 : Bibliographic catalogue of Cephus tabidus. (39) [AsHMEAD, W. H.] In Smith, J. B. 1900. INSECTS OF NEW JERSEY. In Ann. Kept. Sta. Bd. Agr., Suppl., 1899. p. 501-615. Page 600: Species listed as Calamenta [Calameuta'^ johnsoni. <40) Ashmead, W. H. 1903. TWO NEW PHYTOPHAGOUS HYMENOPTERA. hi Can. Ent., V. 35, p. 233. Page 600 : Species described as new under name of Calmeuta johnsoni, from Riverton, New Jersey. BLACK GRAIN-STEM SAWFLY. 17 (41) KoNOW, Fe. W. 1904. sYTEMATiscHE ZTJSAMMENSTELLUNG . . . Clialastogastra. In Zeitschr. Hym. Dipt., v. 4, p, 225-288. Page 256 : Description and synonymy of Trachelus tabidus. Found in South and Middle Europe, North Africa, and Asia Minor. (42) 1905. HYMENOPTERA : FAMILY LYDiDAE. lii Wystman, P., Genera In- sectorum, Fasc. 27. Page 20 : Synonymy of Trachelus tabidus. Found in South and Middle Europe, North Africa, and Asia Minor. (43) RuDOw, F. 1909. LEBENSWEisE j)ER HOLZWESPEN, siKiciDAE. Ill Intern. Ent. Zeitschr., Jalirg. 3, p. 123-124. Page 124 : Record of Trachelus tabidus from South Europe. Also records of host plants which are doubtful. (44) KOKUJEV, NiKITA. 1910. SUE LA DISTEIBUTION DES EEPEESENTANTS DE LA SOUS-FAMILLE DES CEPHINI KONOW ... EN EUSSIE ET LA DESCRIPTION DES ESPfeCES NOUVELLES. In Rev. Russ. Ent., v. 10, p. 127-139. Page 137 : Trachelus tabidus. (45) [MacGilliveay, A. D.] In Smith, J. B. 1910. EEPORT OF THE INSECTS OF NEW JEBSEY. hi Ann. Rept. N. J. Sta. Mus., 1909. Page 595 : Synonymizes Calamenta [Calameuta] johnsoni and repeats record of occurrence in Riverton, New Jersey. (46) Kholodkovsky, N. 1912. A COUESE IN entomology THEORETICAL AND PEACTICAL. PCtrO- grad. V. 2. Page 472 : Trachelus tabidus. (47) KuLAGiN, N. M. 1913. THE PRINCIFAL INSECT PESTS OF FIELD CROPS IN EUROPEAN RUSSIA FOR THE LAST TWENTY YEAEs. In Yearbook Dept. Agr. Centr. Bd. Land Admin, and Agr. Petrogi-ad. v. 6, p. 585-638. (Rev. Appl. Ent., ser. A, v. 2, 1914, p. 201-202.) Records Trachelus tabidus from 26 governments ; more frequent and injurious in southern governments. Damage about 14 to 20 per cent, sometimes more. (48) KURDJUMOV, N. V. 1913. THE MOEE IMPOETANT insects INJUEIOUS to GRAIN CROPS IN MID- DLE AND SOUTH EussiA. In Studies from Poltava Agr. Exp. Sta., no. 17, Dept. Agr. Ent., no. 6, 119 pp. (Rev. Appl. Ent., ser. A, V. 2, 1914, p. 170-173. ) Trachelus tabidus from South Russia on rye. (49) SHT^HEG()LE^, I. M. 1913. EEPOET ON INJURIOUS INSECTS AND DISEASES OF PLANTS IN THE GOVEENMENT OF TAUEIDA DUEING THE YEAE 1912. In Rept. Ass't. Ent. Of Govt. Of Taurida, Simferopol, p. 24-56. (Rev. Appl. Ent., ser. A, v. 1, 1913, p. 357. ) Records Cephus tabidus as having attacked 15 to 20 per cent of the wheat crop. 18 BULLETIN 834, U. S. DEPARTMENT OF AGRICULTURE. (50) UVAROV, B. P. 1914. EEPOET OF THE ENTOMOLOGICAL BUREAU OF STAVKOPOL FOE 1913. In Dept. Agr. Centr. Bd. Laud Admin, and Agr. Petrograd. 86 p. (Rev. Appl. Ent., ser. A, v. 3, p. 44-47.) Records Trachelus tabidus injuring wheat ; not determined if more injurious than Cephus pygmaeus as differences in larvse are still unknown. (51) CORET, E. N. 1914. INSECT PESTS OF 1914. In Kept. Md. Sta. Hort. Soc, v. 17, p. 104-112. Page 106 : Records Cephus pygmaeus injuring wheat at Germantown, Md. (Undoubtedly a misidentification of Trachelus tabidus.) (52) BoBODiN, D. 1915. CEPHUS PYGMAEUS L., ON THE WESTEEN EXPERIMENTAL FIELD OF THE AGRICULTURAL EXPERIMENT STATION OF STAVROPOL, CAUCASUS. In Husbandry, Kiev, p. 876-883. (Rev. Appl. Ent., ser. A, v. 4, 1916, p. 21.) Records Trachelus tatidus infesting stubbles of barley, and winter and spring sown wheat in Stavropol. (53) ZOLOTAEEVSKY, B. N. 1915. PRELIMINARY REPORT ON THE WOEK ON ENTOMOLOGY IN 1914. In Stavropol-Caucasian Agr. Exp. Sta., Stavropol. 12 p. (Rev. Appl. Ent., ser. A. v. 3, 1915, p. 479-480.) Records Trachelus tabidus as appearing later than Cephus pygmaeus, and chiefly found on summer-sown crops. (54) UvAEOV, B. P., & Glazunov, V. A. 1916. REPORT ON THE WORK OF THE ENTOMOLOGICAL BUREAU OF STAVROPOL FOR 1914. In Dept. Agr.. Ministry of Agr., Petrograd, p. 13-54. (Rev. Appl. Ent., ser. A, v. 4, 1916, p. 460.) Records large outbreak of Trachelus tabidus in 1914, with large numbers of parasites. (55) MacGillivray, A. D. 1917. HYMEN OPTERA OF CONNECTICUT. Part 3. Tenthredinoidea. In Bull. 22, Sta. Geol. & Nat. Hist. Survey, Conn. Page 174 : Description of Trachelus tabidus. (56) Howard, L. O. 1918. REPORT OF THE ENTOMOLOGIST. In U. S. Dept. Agr. Annual Repts. 24 p. Page 7 : Record of Trachelus tabidus injuring winter wheat in Northern Virginia, and throughout Maryland and Pennsylvania. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCtTRED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 5 CENTS PER COPY V UNITED STATES DEPARTMENT OF AGRICULTURE O^^f^'^^JT- BULLETIN No. 837 Contribution from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. June 3, 1920 CONTROL OF THE GRAPE-BERRY MOTH IN NORTHERN OHIO. By H. G. Ingeeson, Scientific Assistant, and G. A. Runnek, Entomological Assistant, Dccidnous Fruit Insect Investigations^ CONTENTS. rage. Introductiou 1 History ia Ohio 2 Northern Oliio conditions affecting infestation '2 Varictnl infestation 4 Seasonal history 4 Relation between seasonal-history data and control measures 5 Natural control of first-brood larv8e_ 6 Control experiments 7 Conclusiona 26 Recommendations 26 INTRODUCTION. The grape-berry moth {Polychrosls vltemia Clem.) has been the most destructive insect pest with which the grape growers of north- ern Ohio have ever had to contend. In an effort to improve the methods of control for this insect, extensive experiments in coopera- tion with the Ohio Agricultural Experiment Station were conducted during the seasons of 1916, 1917, and 1918 in northern Ohio. The results of these experiments and the recommendations based thereon are contained in this paper, together with observations made durirg the investigation. Only such life-history data are presented as are necessary for the understanding of the control experiments. The complete life-history data will be presented in a later paper. 1 This investigation was conducted under the direction of Dr. A. L. Quaintance, En- tomologist in Charge of Deciduous Fruit Insect Investigations of the Bureau of Ento- mology. The senior author, assisted by E. R. Selkregg, then field assistant in the Bureau of Entomology, conducted the work during the season of 1916. Much credit is due Dwight Isely of the Bureau of Entomology for his suggestions on grape-insect control, based on similar investigations in the Chautauqua-Erie grape belt of Tennsylvania. The results of Mr. I.sely"s investigations are published in United States Department of Agri- culture Bulletin No. 550. The authors wish to express their appreciation to Trof. H. A. Gossard, entomologist of the Ohio Agricultural Experiment Station, for his help in many ways. To the many grape growers who have cooperated most willingly the authors express their thanks. 147842°— 20— Bull. 837 1 2 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. HISTORY IN OHIO. The grape-berry moth was destructive in Ohio as early as 1869, according to Goodwin.^ This was shortly after grape production became an extensive industry in the State. In 1881 the insect is re- corded as having been especially destructive on the islands in Lake Erie.^ Injury in Ohio was extensive again in 1905 and 190G,but later decreased and in 1909 and 1910 the berry moth caused comparatively little damage ^' -. The infestation became severe again in 1913 and 1914 and reached its height in 1915, when in some local sections as much as two-thirds of the entire crop was ruined. The infestation continued high in 1916, the jSrst season of the investigations here re- ported, and was but slightly less in 1917, Due to a cold autumn, however, the commercial damage was much less in 1917 than in any one of the four years preceding. In 1918, the last year of these in- vestigations, the infestation about Cleveland was of no commercial importance, but in the section about Sandusky the loss was heavy in many unsprayed Catawba vineyards. NORTHERN OHIO CONDITIONS AFFECTING INFESTATION. The grape-berry moth has been a more general pest in the north- ern Ohio section than in the commercial grape sections of New York, Pennsylvania, or Michigan. This statement is based on published reports ^' ■* and on observations made by the senior author during the seasons of 191-1-1918 inclusive. Four principal factors have brought about this condition: The varieties grown, the cultural practices, the method of harvest, and the training system. VARIETIES GROWN. The Catawba variety predominates in the grape section about Sandusky and on the neighboring Lake Erie islands. Due to its late harvest this variety offers ideal conditions for the second-brood larvaj to mature and to reach winter quarters. In all the experi- mental W'Ork conducted the Catawba variety has been uniformly in- fested more heavily than the Concord, which is the predominating variety in the Chautauqua-Erie belt of Pennsylvania and New York and in Michigan sections. CULTURAL PRACTICES, Late in the fall, after grape harvest, a majority of the vineyards are " plowed on." This operation consists in beginning next to the 1 Goodwin, W. II. The grape-berry worm (Polychrosis viteana Clemens). Ohio Agr. Exp. Sta. Bui. 293, p. 259-307 (20 pi. on p. 288-307). 1916. - Gossard, II. A., and Houser, J. S. The grape-berry worm. Ohio Agr. Exp. Sta. Cire. 63. 16 p., fig. 1906. ' Johnson, Fred, and Ilammar, A. G. The gi'ape-berry moth. U. S. Dept. Agr. Bur. Ent. Bui. 116, Pt. II, p. l.'j-71, fig. 4-22, pi. 4-S. 1912. * Isoly, Dwight. Control of the grape-berry moth in the Erio-Chautauqua grape belt. U. S. Dept. Agr. Bui. 550. 44 p., 9 fig., 6 pi. 1917. CONTROL OF GEAPE-BERKY MOTH. 6 vines and plowing three successive furrows of soil toward the vineis. Thus all leaves and trash in the vineyard are covered with from 3 to 5 inches; of soil and ideal winter protection is afforded the hibernating pupae which are in cocoons in the old grape leaves (PI. II, fig. 2). In the spring before time for moth emergence this soil is worked aAvay from the vines. In the Sandusky region it is plowed away and in grape sections near Cleveland it is removed with a disk or worked away with a shovel cultivator. This cul- tivation breaks the crust formed in the winter and in many cases turns to the surface the pupae (PL I, fig. 3) that were plowed under the previous fall. This practice of covering the pupae for the winter and then uncovering them early in the spring protects them from the extreme winter and allow^s the moths to emerge in the spring. Pupae of the berry moth, kept in the insectary yard at Sandusky under conditions similar to those described for the vineyards, lived through the winters of 1916-17 and 1917-18. In the spring of 1917 the emergence Avas 20 per cent and in 1918, after an unusually severe winter, it was 26 per cent. Comparative data are not at hand for the same winters with pupae exposed as they would be in a vineyard plowed before grape harvest and then left until spring. In experi- mental work reported by Isely,' however, it was found that subsequent emergence from cocoons left through the winter of 1915-16 under exposed conditions in the vineyard was but 5 per cent as com- pared with 30 per cent emergence where the cocoons were covered by 2 inches of earth and then uncovered before time for emergence in the spring. Since late plowing away in the spring is objectionable in northern Ohio from a horticultural standpoint, the writers recommend that when cultivation is completed in July the vineyards, w^henever pos- sible, be placed in final cultural condition for the winter and then that they be left in that condition until the next spring. The only objection to this practice is the excessive growth of weeds, which in- terferes with harvesting. This can be overcome by seeding a cover crop at the completion of cultivation. METHOD OF HARVEST. A large part of all the grapes in these sections and practically all of the Catawba variety w^ere formerly sold for wine making. Since no particular packing is required for this market, all sorting is done in the vineyards by the pickers. Wormy berries are cut or shaken out of the clusters and allowed to fall to the ground and to remain in the vineyard. It appears that any other method of dis- posing of the infested grapes would be more costly in labor than would be warranted now that satisfactory control may be secured by spraying. 1 Isely, Dwight, op. cit. 4 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. TRAINING SYSTEM. The "■ fan " system of grape training (PL III, fig. 1) which is used consistently in northern Ohio with the Catawba variety and a modi- fication of which is used with Concords, is not generally practiced in any of the other commercial grape sections of the country. This "fan " system consists in securing the bearing canes from the old vine head betw^een the ground and the first wire and tying them up obliquely to the first and second wires, forming a V open at the top. Although two canes are the rule with the Catawba variety, when the thrift of the vine allows of more than two the additional canes are also carried up obliquely, completing the fan from which the system takes its name. As the young shoots bearing the clusters grow to a sufficient length they are tied up vertically to the middle and top wires. An effort is made to have these shoots spread, but to economize labor in tying they are often bunched 2 to 4 in a place. This system of training spreads the grape clusters all through the vine from the ground to the top wire and covers them almost com- pletely with foliage and shoots (PI. Ill, fig. 2). These conditions explain in part the failure to cover the grape clusters when any set- nozzle method of spraying is used, particularly when the spraying is done late in the season and considerable vine growth has been at- tained. VARIETAL INFESTATION. Several commercial varieties -of grapes are present in northern Ohio, affording opportunity for observation on the relative infesta- tion of the different varieties by the grape-berr}- moth. A list of varieties observed, beginning with the most heavily infested and ending Avith the least, is as follows: Shride, Elvira,,Clinton, Reisling, Catawba, Norton, Niagara, Delaware, Agawam, Ives, Concord, Wor- den, and Moores Early. In general it seems that the early-blooming varieties like the Shride and Clinton become heavily infested with the first-brood larvae, and late-harvested varieties like Catawbas and Nortons be- come heavily infested with second-brood larvae. SEASONAL HISTORY. The grape-berry moth completes one life cycle and a part of an- other each season. This insect is injurious only in the larval stage. There are two broods of worms or larvae every season (PI. I, fig. 1), the second much more numerous and destructive than the first (PI. II, fig. 1). Since an understanding of the main points in the life history of the insect is necessary for the best application of control methods, a brief summary will be given. CONTROL OF GRAPE-BERRY MOTH. 5 The winter is spent in the pupal stage in cocoons (PI. I, fig. 2) which the Larvae spin in grape leaves the previous fall. These leaves are the ones that fall early and become soft and sodden on the ground (PI. II, fig. 2) and remain under the trellis during the winter. In the spring, previous to and during grape bloom, moths (PI. I, figs. 4, 5) begin to emerge from the overwintering pupae. This emergence gradually increases and continues at a high point for about 3 weeks. The moths begin to deposit eggs on the young grapes about 4 days after emergence and the eggs hatch in from 4 to 6 days. This first brood of larva3 or worms usually is not seriously destructive, though first-brood infestation amounting to as much as 30 to 35 per cent has been observed. The average length of the feeding period of this brood of larva? is 23 days. When mature the larvae migrate to grape leaves on the vines and spin their cocoons on them. From the cocoons moths emerge in about 13 days and begin laying eggs about 4 days later. The eggs of this second brood are placed on the nearly full- grown grapes and are easily found where the infestation is heavy. Before the eggs hatch they appear as creamy-white raised dots on the green grape berries, but after the larvae leave the eggs the egg- shells appear as glistening white spots. This brood of eggs hatches in from 4 to 6 days and it is the resultant brood of larvae that, if allowed to develop, does the greatest damage to the grape crop. (PI. II, fig. 1.) The larvae of this brood feed for a long period and usually leave the grapes just before harvest. They spin down to the ground and make their winter cocoons on old decayed grape leaves under the trellis. In the case of a cold fall many larvae do not leave the grapes but are harvested with the grape crop. This condition prevailed in the fall of 1917 to an unusual degree and the result was a lighter infestation in 1918. RELATION BETWEEN SEASONAL-HISTORY DATA AND CONTROL MEASURES. The control experiments recorded in this bulletin are based on ex- tensive field observations and on life-history studies conducted each season. The data shown in diagram form in figure 1 are sum- marized from the complete seasonal-history data. In determining the hatching periods of the larvae 4 days are allowed from the emergence of the moths to the deposition of eggs and 6 days for in- cubation of the eggs. These are average figures from many obser- vations extending over several seasons. It is seen in figure 1 that in 1916 and 1917 a few larvae had hatched before Concord grapes began to bloom and in 1918 that the dates of first hatching and beginning of bloom are coincident. In each sea- son the first-brood larvae were hatching in large numbers for about 6 BULLETIlSr 837, U. S. DEPARTMENT OF AGRICULTURE. 3 weeks. It is important to note that the rise in tlie early part of the hatching is abrupt and the subsidence of hatching more gradual. It has been the opinion of other writers that the largest part of the second-brood larvte hatch within a shorter period of time than the first brood. The rearing records here illustrated do not support that belief but show the hatching periods to be of about equal length. NATURAL CONTROL OF FIRST-BROOD LARViE. It was observed that the grape berries infested by first-brood larva? dropped readily from the vines when touched. It was thought that if these infested berries dropped in any great numbers at any particular time some cultural method such as covering these berries with soil might aid in the control of the insect. To determine this /o so /o so /O SO /o so 3o /" ^N, / <5> / > \c / V ^^^y /- ■ ■ \, / ^- N / V J. V ^ ^^^ / \ -. — y \ ^.^0>^ Fig. 1. — Diagram showing relation between dates of spray application and periods during which the grape-berry moth larvae were hatching for the seasons 1916, 1917, and 1918 at Sandusky, Ohio. point the following experiment was undertaken in 1916: Wooden frames were made, G feet long, 30 inches wide, and 6 inches deep with cheesecloth stretched on the bottoms. These trays fitted be- tween the vines directly under the trellis and were placed in six different locations in the vineyard, under vines heavily infested with first-brood larvse. Fresh leaves were supplied in the trays for the cocooning of any larva^ that might drop. The trays were put in place July 5 when first-brood infestation was about at its height and were left until August 15, when practically all first-brood larvae had left the grapes. These trays were examined every three days. Practically no grapes dropped from the vines and not a single larva was taken throughout that period. From these negative results it is concluded that practically no natural control occurs from the drop- ping of grapes infested by first-brood larvae. CONTROL OF GRAPE-BERRY MOTH. 7 CONTROL EXPERIMENTS. STATUS OF SPRAY PRACTICE FOR GRAPE-BERRY MOTH CONTROL. When these investigations were undertaken the following prin- cipal facts were known about spraying for the control of the grape- berr}' moth: First, satisfactory control had not been effected by the use of any sj^stem of set-nozzle spraying, particularly in thrifty vineyards where foliage growth was heavy. Second, satisfactory control had been effected in Ohio^ by using the trailer method of spraying at the time of the hatchings of second-brood larvae, usually in early August. This practice, however, left a heavy residue of spray material on the fruit at harvest time, which tended to exclude such fruit from a basket market. Third, two spray applications by the trailer method, the last when the grapes first touched in the clusters, had given satisfactory control on the Concord variety in the Chautauqua-Erie belt in 1915. This practice was to be thoroughly tried in northern Ohio on Concords and Catawbas. SCOPE OF EXPERIMENTS. From this summary of the knowledge available it appeared that the investigations should deal with three main points: (1) Time and number of spraj^ applications, (2) chemicals used in spray materials, (3) spray residues left at harvest time. In studying these factors spraying experiments were conducted by the writers in 6 vineyards in 1916, in 9 in 1917, and in 15 in 1918, a total of 30 vineyards. These vineyards were selected for the op- portunity they offered for the advantageous study of any one or more of the important features enumerated above. Since little would be gained by considering each vineyard separately it has seemed desirable to assemble in Tables I, II, and III the data relating to the different vineyards and to bring together in similar form in Table IV the results of the experiments. TIME OF SPRAY APPLICATIONS. Former experiments " indicated that a spray application directly after grape blooming was important for the control of both grape rootworm beetles and grape-berry moth larvae. In the Sandusky and island sections of Ohio a spray application following grape bloom is usually made for the control of downy mildew, PlasmoplwTa I'^lticolaj particularly on Catawba and Delaware varieties. This application directly following grape bloom was considered as the first spray in all of the experiments in which it was included. The second spray 1 Goodwin, W. H., op. cit. = Goodwin, W. H., op. cit. John.son, Pred, and Hammer, A. G., op. cit. Isely, Dwight, op. cit. 8 BULLETIN" 837, U. S. DEPARTMENT OF AGRICULTURE. was applied when the grapes touched in the clusters, but before the clusters were tight enough to prevent the spray material from being driven between the grapes. This stage of grape growth usually oc- curs from 3 to 4 weeks after bloom. This second spraying was designed to kill the late hatching first-brood larvae and to remain on the grapes to be eifective when the second-brood larvae hatched. The third spraying was timed in each case to precede immediately the hatching period of the majority of second-brood larvae. METHOD OP APPLICATION. All spraying was done by the hand or trailer method in which 2 hose lines of from 20 to 50 feet trail behind the sprayer, and the spray material is delivered through short spray rods and angle noz- zles, directed by hand as in tree spraying. Variations in this method will be discussed later. Sufficient pressure was maintained to drive the spray well into the clusters, but the amount of pressure varied from 125 to 225 pounds in different vineyards. The best pressure to maintain will vary somewhat with the vineyards, but the writers believe that from 175 to 225 pounds usually will be found most efficient. Nozzles set at an angle are absolutely necessary for efficient work, and it was found that nozzles set at angles of 45° allowed more freedom of handling than those set at 90°. A nozzle aperture of -/^j-inch was most commonly used, but the most efficient size was found to vary with the vineyard and other local conditions. WEATHER CONDITIONS 1 AFFECTING SPRAY RESULTS. The season of 1916 was about normal in all respects except for an unusually dry period during July and August which was favorable for spraying and for spra}^ material adhering. These same condi- tions were likewise favorable for the development of an unusually large second brood of worms. September and October were warm and dry, conditions also favorable to extensive berry moth injury as shown in the uniformly heavy infestation in the checks (Table I V) . In 1917 conditions were decidedly unfavorable for spraying opera- tions. Both the first and second applications were interfered with by rain and closely followed by showers of varying intensities. In July the total rainfall was but 0.46 inch, but this came between the first and second spray applications. The maturing of first-brood larvae was favored by an exceedingl}' hot and di'j' period from July 28 to August 6 and a subsequent heavy hatching of second-brood larvae followed. September was 3.3° below normal in temperature and slightly below in precipitation, while October was 8.3° below normal with 3.79 inches of rainfall above normal. These unfavorable ' Weather records from the U. S. Weather Bureau Station at Sandusky, Ohio. Bui. 837, U. S. Dept. of Agriculture. Plate I. ^iS The Grape-berry Moth (Polychrosis viteana). Fig 1 - Larva. Fig. 2.-Pupa (ventral aspect) in cocoon. Fig. 3.-Pupa (dorsal aspect). Figs. 4, 5.— Adult. All greatly enlarged. Bui. 837, U. S. Dept. of Agriculture. PLATE II, Bui. 837, U. S. Dept. of Agriculture. Plate III. The "Fan" System of Grape Training. Fig. 1.— Northern Ohio vineyard trained according to the fan system. Fig. 2.— Grapevine showing fan system of training with grape clusters scattered from near the ground to the top wire. 1 CONTROL OF GRAPE-BERRY MOTH. 9 conditions in September and October retarded the development of second-brood larva? and counteracted the previous favorable condi- tions. The season of 1918 opened unusually early and continued favor- able for all growth processes throughout the season. Spraying was but little interfered with and no unusual weather conditions pre- vailed that affected the spraying results. SPRAYING EXPERIMENTS IN 1916. Table I. — Vineyards used for spraying experiments in northern Ohio, 1916. Vine- yard Vineyard owner and location. Varieties. Dates of spray applications. Esti- mated infes- tation 1915. Gallons of spray ma- terial per acre. No. First. Second. Third. First. Second. Third. 1 2 3 4 5 Roland Brown, Kelleys Island. 0. W. Brown, Kelleys Island. Charles Duggan, Put- in-Bay. W. R. Huntington, Put- in-Bay. E. Mantv, Venice Catawbas . ....do ....do Catawbas, Concords. ....do July 5 July 3 June 30 July 1 June 29 June 27, 28. July 21 July 20 July 17 July 19 July 12 July 11, 12. Aug. 9 ....do... Aug. 7 Aug. 8 Aug. 3 Aug. 2 Perct. 70 85 75 80 90 70 173 120 112 105 128 90 192 202 176 160 304 236 192 200 112 112 150 C John Schonhart, Venice . ....do 230 Experiments were conducted in six vineyards in the Sandusky and island sections, as shown in Table I. In all of these experiments the arsenicals w^ere applied in Bordeaux either 3-3-50 or 2-3-50 strength. Laundry soap at the rate of 2 pounds to 50 gallons was used for the first spray application in all of the vineyards and for the second spray in vineyards Nos. 5 and 6. It became apparent dur- ing the second spray application that resin fish-oil soap possessed better spreading qualities than laundry soap and so it was used in all the other spray applications at the rate of 1 pound to 50 gallons. Previous to bloom all of the vineyards received an application of Bordeaux alone for the control of downy mildew. The strength of arsenicals in the berry-moth sprays was varied in the different vine- yards as shown in Table VI. 147842°— 20— Bull. 837 2 10 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. SPRAYING EXPERIMENTS IN 1917. Table II. — Vineyards used for spraying experiments in northern Ohio, 1911. Vineyard o^vner and location. Varieties. Dates of spray applica- tions. Esti- mated infes- tation 1916. Con- trol of grape- berry moth in 1916. Gallons of spray material per acre. Vine- yard No. First. Second. Third. Spray applications. First. Sec- ond. Third. 1 Becker Wine Co., Kellys Island. O.W. Brown, Kel- leys Island. Paul Cooley, Do- ver Center. Fred Foye, Put-in- Bay. George Lewis, Bay Village. John Schonhart, Venice. T. W. Wearsch, Avon Lake. Catawbas . ....do Concords . . Catawbas . Concords, Ives. Catawbas, Concords. Concords, Ives. July 14 July 12 July 12, 13. July 11, July 5-9 July 3, 4 July 10 Aug. 3, 4 Aug. 2 July 26 July 31, Aug. 1. July 23- 26. July 20- 25. July 27 Perct. 25 50 40 90 40 20 90 (3) (') (3) (3) 150 165 250 150 200 150 120 193 178 300 233 275 312 100 2 3 ■1 Aug. 18 300 5 6 7 Aug. 17 Aug. 14 Aug. 17 200 280 160 1 One spray ^^ith trailers. 2 Two sprays with trailers but operators riding. 3 No spray. * Two sprays with trailers. As seen in Table II the experiments were conducted in seven vine- yards, four in the Sandusky and island sections and three in the Dover and Avon sections just west of Cleveland, Ohio. The ex- periments were so placed as to include local variations in the dif- ferent grape sections, such as weather, cultural practices, varieties, and markets. All the arsenicals were applied in Bordeaux 2-3-50 with resin fish-oil soap added at the rate of 1 pound to 50 gallons. The following single exception was made : In vineyard No. 5 copper sulphate was omitted from the third spray application and laundry soap, 2 pounds to 50 gallons, was substituted for resin fish-oil soap. SPRAYING EXPERIMENTS IN 1918. Table III. — Vineyards used for spraying experiments i)i Northern Ohio, 1918. Vineyard owner ' and location. Varieties. Date sof spray applica- tions. Esti- mated infes- tation 1917. Con- trol of grape- berry moth in 1917. Gallons of spray material per acre. Vine- yard No. First. Second. Third. Spray applications. Fust. Sec- ond. Third. 1 2 CD. Powell, Ver- milion. O. W.Brown, Kel- leys Island. T. W. Wearsch, Avon Lake. Ernest Dimning, Avon Lake. V. Doller, l>ut-in- Bay. Ives Catawbas . Concords, Ives. Concords.. Catawbas . Jime 13, 14. June 29 Jime 14, 15. Jime 19 July 2-3 July 9 July 18 July 11, 12. July 16 Omitted. Aug. 5 Perct. 15 15 2 50 75 0) (') 100 200 (=) 150 342 (2) (2) (2) 3 4 5 1 Two sprays with trailers. 2 No record. 8 No spray. I CONTROL OF GRAPE-BERRY MOTH. 11 The experiments were extended in 1918 to include work in 15 vine- yards, but as infestation was not sufficiently heavy for satisfactory comparisons in all the vineyards only the 5 showing the heaviest infestations are included in Tables III and IV. The arsenicals were applied in Bordeaux. 2-2-50 in vineyards Nos. 1, 2, and 5. In vineyards Nos. 3 and 4 copper sulphate was omitted at the request of the owners. Stone lime 2 pounds to 50 gallons was retained to care for any free arsenic in the arsenicals. Resin fish-oil soap at the rate of 1 pound to 50 gallons was used uniformly through- out the experiments. METHOD OF RECORDING RESULTS OF SPRAYING EXPERIMENTS. It had been learned in earlier work ^ that results based on weights of harvested fruit were misleading, owing to the varying thrift of vine- yards, time of harvest, weather conditions affecting the development of worms, etc. The weight method, therefore, was abandoned in favor of the count method. This consists in selecting a representa- tive number of vines in each sprayed plat and in each check, har- vesting all the fruit from these vines, counting the clusters, then the clusters containing wormy berries, then removing the wormy grapes and counting them. To ascertain the average number of grapes per cluster, 100 representative clusters were taken in each vineyard and all the grapes counted. The number of clusters in each plat was then multiiDlied by the average number of grapes per cluster to give the total number of grapes examined in each plat. In all control work on the grape-berry moth, the unevenness of in- festation within a vineyard has made experimental results difficult to interpret. This uneven infestation prevailed throughout these in- vestigations but was cared for whenever possible by placing checks across the control plats and reading the results on the control plat the second post-length away from the checks. While the plan does not entirely overcome the difficulty, the writers feel that the averages from several vineyards closely approximate actual conditions. In all cases the fruit from at least 10 vines was examined and when possible the examinations included all the fruit from 20 to 25 vines. Exceptions to this occurred only when there were less than 10 vines of a particular variety in a plat. First-brood counts were made in some instances, but since they add little to the final results they are omitted from the tables. * Johnson, Fred, and Hammar, A. G., op. cit. ; Isely, Dwight, op. cit. 12 BULLETIISr 837, U. S. DEPARTMENT OF AGRICULTURE. S g to o y, rS o o H Si < -4j U 00 s '^ fri s -"J <3 -d ^ ^s^ A 1 I 03 b . V, o • >-o n Z fci ^ CO <1 < 03 8 o u W CO S t cl H '-. =0 OQ O P4 1 ft C3 . tH Oi f^ M S a % S, > S < t^ tM CO PQ ^ S -§ |j ^ 1 s « 5> b O >5 tH BS p Eh ox) .3^ > .9^ Of O BY. > > > o 3 0:3 o ca 1^ •3I09qo jnaoBfpv •l^Id pa.iBJdg 00 1/5 00 r* r* b* Oi M 00 CO w ■"j'oo OJ W as CO l08qo ^uaoEfpv •^nid pajiBJdg Jioaqo ^uaocrpv ■j^ld paiBJdg •qai'Bqoi mS -9q ffiAJBi pooj'q - p u 0 0 9 s naqAV ■sJOisTqo aqi uj q 0 n 0 1 sadBjS uaqAV •sn^j nioojq •IBIJ9?BH ■japAiod peaijo atBuasjy •snoj:)BonddBjo jaqmuM OC-ICOcOCCt^COCOOifM COCOOit^ClcOfMOOCOCO •^ CO t^ ^ w 01 o (>i -^ ^ co-^t^-^r^wt^iocO"^ rH-^t^^H^COOCOOO^ rJ CONTROL OF GRAPE-BERRY MOTH. 13 gS Ol l>- 00 OS M< t^ 04 , OO lOOOcDCS lOIOcD -^^ toio CO lO »-H CO "^ ^ ut .-1 to CO 00 CO CO -^ r^ (M o r^ ■* CO •<*< lO O CC"-" CD -^ C4 CO iC t^ Oi Cs TJ*C^ T-nr Tt CO ^tc o OOmQw uu wtJOOC mO X M _/:^ WN X M -£ M M X M K M s s g >OiO 5 S .^n Hn HC^ HC-IHCI Hc< HN M t-l '"' f-t w^ N e 1 1 i • ^ •1 • fe m CI 1 8 § XX K XXX X XX X : X lO lO lO lO lO lO Hlf» r«4<-«r4 i-H p^M P*N IM oq CO (N '-' 14 BULLETIN 837, TT. R. DEPARTMENT OF AGRTCULTUEE. In Table IV are brought together all of the data bearing on the time and number of sj)ray applications, arranged by vineyards and varieties. This table provides for a comparison of the plats within each vineyard by reading from top to bottom, as well as a com- parison of the plats receiving similar spray treatment in the different vineyards, by reading across. While the comparison between plats within a vineyard is relatively consistent, considerable variations exist between vineyards. A study of the column of averages shows satisfactory commercial control to have been effected in all plats which received either two or three spray applications with the excep- tions of the first two plats. These two plats illustrate the necessity of timeliness of spraying and adhesiveness of spray material, since in 1917 the more timely spray treatment, and the use of resin fish-oil soap throughout, reduced the average infestation from 19 per cent in 1916 to 5 per cent in 1917. RELATIVE EFFICIENCY OF DIFFERENT TIMES AND NUMBERS OF SPRAY APPLICATIONS. * Table V. — Summarised results from Table IV — relative efficiency of different times amd nnmhers of spray applications, 1916, 1917, and 1918 — Arsenicals applied in Bordeaux mixture and soap solution^ 8 01 S s 11 1 ffi a) la Percentage of grape berries infested, aver- a) SI s5 1 ages of all experiments, 1916, 1917, 1918. 1 Catawbas. Concords. Ives. a a 'So 0) 01 ^1 a i i ,^5 8 M 8 «3 "o o ■^■a ^g 13 o 5s p. T3 C3 ■s p. o x> « TS I 2 C3 o ^ ^ g t» S 0 3 M 0) ID a 3 a 3 s 03 2 p< ■■5 2 '■V ^ f^ M e < r^ 15 ;z; QQ < w M < 2 X X 2i 1916,1917 9 11 11.35 79.22 3.04 48.20 2.11 39.67 2 . X X 24 1916 3 4 18.99 83.98 2.31 22.67 2 X X li 1917,1918 11 14 4.67 64.92 .82 26.00 1.53 16.85 2 X X li 1917,1918 5 6 l.U 35.64 2.07 30.14 2 X X X X 2i 2i 1916 1916 3 6 4 8 5.78 6.86 79.27 79.98 2.66 1.32 22.67 66.26 3 X 3 X X X U 1917,1918 5 5 .41 42.00 1.21 52.96 1.31 17.92 1... . X 2i 1 1916,1917 1917,1918 1918 6 4 2 7 4 2 16.90 79.77 6.21 13.30 31.91 26.40 7.23 8.75 2.08 39.67 1 X X 9.89 2.. . X 7.39 62.36 6.93 1 X n 1918 2 2 25.67 82.82 4.20 10.04 A comparison of the com,binations of the first and second sprays with first and third sprays shows little choice between them as far as berry-moth control is concerned. Since the combination of the first and second sprays leaves the fruit practically free of all spray residue at harvest time and since the second spray is more easily ap- plied than the third because of lighter grape foliage, this combina- tion of the first and second application is preferred by the writers. It is important to know, however, that if for any reason the second spray can not be made, the third may be applied and will give about CONTROL OF GRAPE-BERRY MOTH. 15 equally as good control. When the third application is made the fruit Avill usually be unfit for basket market because of excessive spray residue. Where a third application is added to the first and second, control is slightly better, 6.8 per cent infestation as compared with 11.3 per cent in 1916 and 0.4 per cent as compared with 4.6 per cent, an aver- age for 1917 and 1918. These differences, however, were not suffi- cient to justify the expense of the third application. ONE-SPRAY METHOD. One spray application by the trailer method at the time the grapes first touched in the clusters gave an average control of 83 per cent on Catawbas and 94 per cent on Concords as compared with 89 per cent control on Catawbas and 97 per cent on Concords when both the first and second sprays were given. This is a good showing for a " one-spray " schedule and this treatment might be the most efficient under some conditions. All results indicate that this one spray may be depended upon to save the crop from ruin by the berry moth. In an effort to eliminate entirely the factor of spray residue on fruit for the basket market the spray application directly after grape bloom was tried alone. Data are insufficient on this treat- ment but indicate a marked effect on the final infestation. This treatment has the advantages of being the most important one for rootworm beetle control and of being timely to prevent black-rot in- fection of the young grapes. It may develop that this method will be practical after the infestation of the moth has been reduced by the use of the two-spray schedule for one or more years. Experiments on this point were conducted in several vineyards in 1918 but ad- jacent checks failed to show sufficient infestation to make results conclusive. CONCLUSIONS FROM EXPERIMENTS. The combinations of the first spray treatment with the second and of the first with the third gave satisfactory control. The third spray added to the first and second increased the effectiveness, but not enough to justify the expense of making the application. The second application alone averaged 83 per cent control and in all cases saved the commercial crop. The first application alone reduced the final infestation appreciably but needs further testing. MATERIALS USED IN SPRAYS. Absenicals. aesenate of lead, commercial powder. Arsenate of lead in powder form was used throughout this work. Since previous infestation had been extremely heavy in the experi- mental vineyards, the powder was used in 1916 at the rate of 2^ 16 BULLETIN" 837, U. S. DEPARTMENT OF AGETCULTURE. pounds to 50 gallons, equivalent to 5 pounds of paste to 50 gallons, as a basis for comparison that year. The 1^-pound rate was used in but two vineyards in 1916. In 1917 the 2^-pound rate was retained in four vineyards and the 1^-pound used in seven. In 1918 compari- son was made between 1 pound and 1^ pounds of arsenate of lead powder. Table VI. — Relative efflciencp of arsenate of lead at the rate of 1, 11, and 2i pounds (po^cder) to 50 gallons liquid^ Spray appli- cations. Percentage of infested grape berries. a Catawba va- riety. Concord va- riety. Ives variety. Poiinds of arsenate lead 3 o . powder to 50 gallons liquid c3 g a V o ft o d ft J ft ■a 4<( a ft o 1 a lO a ^ Xi t>> >, t>> ^ (0 (U ea C3 S3 e? ■S 3 3 ;H f3, "^ K -o n, ■a CO ^ Z ;z; > m ■< M < 02 ■< SEASON 1916. {i X 2 3 4 6 0.93 11.90 81.12 70.26 li 5.56 52.83 2 3 6.42 75.69 5.56 62.83 {I X X 2 3 4 6 3.25 3.34 64.52 43. ,32 10.41 64.39 2 3 10.41 64.39 3.30 53.92 SEASON 1917. f X X X X X X X X X X X X X X 1 2 3 4 5 6 7 2.03 8.86 64.09 78. 17 0.47 61.61 IJ 5.84 63.03 .84 .18 2.66 32.11 20.60 88.03 0.43 2.94 20 14 1.48 42.00 19.59 Averages 7 10 4.55 61.82 1.04 60.59 1.68 19.86 li X X X X 1 2 4 6 .67 6.47 7.72 64.09 79.31 79.63 *J .48 30.14 2.11 39.67 4 5 4.95 74.34 .48 30.14 2.11 39.67 SEASON 1918. f X X X X X X X X 1 2 3 4 0.86 7.85 4.78 68.02 13 0.41 1.00 2. .54 6.28 4 4 4.78 68.02 .71 4.41 .86 7.85 {i X X 1 2 2.08 6.93 7.39 62.36 Averages 2 2 7.39 62.36 2.08 6 93 I All plats received Bordeaux 2-3-50 and resin soap 1 pound to 50 gallons in 1917 and 1918. Same in 1916, except laundry soap 2 pounds to 50 gallons, first application; and second application vineyards Nos. 5 and 6. 3ul. 837, U. S. Dept. of Agriculture. Plate IV. I CONTROL OF GRAPE-BERRY MOTH. 17 The differences in the control effected by the various strengths of arsenate of lead were slight, as shown in Table VI. When compar- ing the averages the differences in the infestation of the adjoining checks should be kept in mind. The tests of 1 pound to 50 gallons have not been sufficient to justify one in drawing conclusions, but 1^ pounds to 50 gallons has proved adequate for control. ARSENATE OF CALCIUM, COMMERCIAL POWDER. Much interest has centered in the comparative merits of arsenate of calcium and arsenate of lead as insecticides. Arsenate of cal- cium has the advantage of being much cheaper than arsenate of lead, but doubt has prevailed as to its adhesive qualities and its effect on foliage. Since an extra spreader and adhesive in the form of resin soap is necessary even with arsenate of lead for spraying grape clusters, and since grape foliage is comparatively hardy to arsenicals, it was thought that arsenate of calcium should have a wide use in grape spraying. Table VII. — Relative efficiency of coiiDiiercial arsenate of calcium and arsenate of lead for control of the grape-berry moth, Sandusky, Ohio, 1917, 1918. Both arsenicals applied in Bordeaux 2-3-50 uith 1 pound of resin soap to each 50 gallons. Spray appli- cations. C3 p. 0 'A 0 •s C3 Percentage of infested grape berries. Arsenical, pounds to 50 gallons. ft i~€ !i B ■a >o 0 CO 3 0 i cl Catawba va- riety. Concord va- riety. Ives variety. ft 2 ft M .§ 0 a ■a < C3 ft >, C3 Ut ft m ii 0 1 0 i < r! ft C3 ft m 0 .a a •i < X X 1917. 1 4 5 6 1918. 1 2 9.46 13.78 64.09 79.63 Arsenatfi of calcium 0.58 .66 32.11 40.60 commercial powder, 42 per cent AsjOs, 1 pound to 50 gallons 1.28 1.04 39.67 6.23 7.22 87.06 Averages 6 7 10.15 76.93 .62 36.35 1.16 22 95 [^ X 1917. 1 4 5 6 1918. 1 2 2.03 5.84 61.09 63.03 .84 .18 32.11 20.60 mercial powder, 30 per per cent AS2O5 2.94 .86 19.59 7.85 4.78 68.02 Averages 6 7 4.22 65.05 .51 26.35 1.90 13.72 Table VII shows the comparison of arsenate of calcium with arsenate of lead in six vineyards and seven plats. Control was 18 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. almost complete for both materials on the Ives and Concord varie- ties. On the Ciitawbas the arsenate of calcium averaged 90 per cent control and the arsenate of lead 96 per cent, but the checks adjacent to the arsenate of calcium plats were 12 per cent more heavily infested than those adjacent to the arsenate of lead plats so the comparisons are very close. These results may indicate that the arsenate of lead adhered slightly longer in the season than the arsenate of calcium. In recording the spray residue on the fruit at harvest time, slightly less was found on the arsenate of calcium j3lats than on corresponding arsenate of lead plats. This feature is an advantage in grape-beiTy moth spraying and is discussed later. No foliage injury that could be attributed to the arsenical occurred on any of the arsenate of calcium plats. Grape spraying experiments were continued in 1919 and foliage injury occurred on all plats of the Ives variety where commercial arsenate of calcium was used at the rate of 1^ pounds to 50 gallons of water with 3 pounds of freshly burned stone lime slaked and added to each 50 gallons of spray solution. This experience indi- cates that the use of arsenate of calcium on the Ives variety is unsafe. ARSENATE OF CALCIUM, HOME-MADE PASTE. Varied success had been reported from the use of home-made arsenate of calcium pastes as sprays for fruit trees. To determine the efficiency of these home-made materials for use in sprays on grapes the following experiments were conducted. Pastes were made according to the following formulas and methods and applied in spray solutions to grapes: (1) Arsenate of soda 4- stoue lime. Sodium arsenate, fused (dry powdei-etl) 60 per cent AsaOe ■ oiinces— 30 Stoue lime do 18 Water do 48 Total do 96 The sodium arsenate was dissolved in the water and the resultant solution used to slake the lime. A smooth paste arsenate of calcium of about 18 per cent AsgOg content resulted. This was decanted 5 times to remove the sodium hydroxid. The resultant paste was used at the rate of 2|- pounds to 50 gallons to be comparable with arsenate of lead (commercial powder 30 per cent AsgOg), 1^ pounds to 50 gallons. (2) Arsenic acid -f- stone lime. Arsenic acid (liquid) 78 per cent AS2O6 ounces 10 Stone lime do 8 Water do 34 Total do 52 CONTROL OF GRAPE-BERRY MOTH. 19 The lime was slaked to a smooth paste with 18 ounces of water. The arsenic acid was diluted with the remaining 16 ounces of water and the diluted acid added to the lime paste a little at a time. The paste was stirred vigorously during the mixing. With each addi- tion of acid the lime had a tendency to granulate, but continued stirring restored the smooth pasty condition. A sample of the final paste was analyzed by the United States Bureau of Chemistry under miscellaneous laboratory No. 24714 as follows: Moisture 69. 7 Total CaO (as received) 15.08 Total AS2O5 (as received) 12.04 Water soluble AS2O5 .02 5 grams samples in 1,000 c. c. CO2 free water ; equivalent to 2 poimds to 50 gallons. Free lime calculated as calcium hydroxid Ca(0H)2 4.4 per cent. Paste made according to this formula was used on grapes in 1918 at the rate of 4 pounds to 50 gallons of water. Grape-berry moth infestation failed to develop in numbers sufficient for the desired comparisons in any one of the four vineyards in which these ma- terials were used. The spreading qualities and effects on grape foliage, however, are important. The arsenicals were applied either in Bordeaux 2-2-50 or with stone lime 2 pounds to 50 gallons added to the mixture. In all cases resin fish-oil soap at the rate of 1 pound to 50 gallons was also added. The paste made from sodium arsenate spread equally as well as the commercial arsenate of cal- cium powder or arsenate of lead powder. The paste made from ar- senic acid failed to spread as well and when dry it was not a smooth, even coating such as is desired. In no case could foliage injury be attributed directly to the use of either of the home-made arsenate of calcium pastes. Spreaders and Adhesives. Because of the partial failure of arsenate of lead and Bordeaux to spread over or " w^et " the individual grapes in the grape clusters, various materials have been added to these to facilitate the spreading process. The qualities desired in such a spreader are (1) quick- wetting power, (2) adhesive power when dry, and (3) that it be easily pre- pared for use. In addition, a material to be suitable must be com- patible with Bordeaux mixture and arsenicals and also be compara- tively cheap. Former investigations ^ had shown that some form of soap was the most practical material for the purpose. When these investiga- tions were undertaken various soaps were recommended by differ- ^ Isely, Dwight, op. cit. 20 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. ent authorities. To determine the most efficient of these, experiments as shown in Table VIII were conducted in 1916. Table VIII. — Relative efficiency of different soaps as spreaders and adhesives, Schonhardt vineyard, Venice, Ohio, 1916. Pounds in 50 gallons. Spray mate- rials com- bined with soaps. Gallons spray mate- rial per plat second appli- cation. In- crease over resin soap. Results in grape-berry moth control. Kind of soap used. Num- ber vines exam- ined. Num- ber clus- ters exam- ined. Num- ber grapes exam- ined. In- fested grapes. Variety. Soft 2 2 1 0) (') 140 100 75 Per ct. 86 33 0 / 1^ 1 10 / 11 1 10 / 12 I 10 / 22 I 34 790 342 608 198 735 311 1,050 1,019 18,960 12, 996 14,592 7,524 17,640 11,918 25,200 30,570 Per ct. 17.04 24., S3 6.78 6.11 8.50 8.37 41.16 73.12 Concords. Laundry Concords. Resin fish-oil Catawbas. Concords. Checks Catawbas. Concords. • Bordeaux 3-3-50, arsenate of lead commercial powder 2^ pounds to 50 gallons. Adjoining grape rows, each row inchiding Concord and Catawba varieties, were sprayed three times during the season, the first ap- plication three to five days after grape bloom, June 27, the second when the grapes touched in the clusters, July 13, and the third at the beginning of the hatching period of second-brood larvae, August 3. All spraying was done by the trailer method with medium disk angle nozzles and at a pressure of 150 pounds. The soaps were used as spreaders in mixtures of Bordeaux 3-3-50 and arsenate of lead powder 2| pounds to 50 gallons. The season was unusually dr.y dur- ing July and August, favoring both adherence of spray materials and the development of the grape-berry moth. The Concords were harvested September 29 and the Catawbas October 10. The soft soap used was a bulk product made especially for use in commercial laundries. This soap dissolved readily in hot water, but when applied to grapes formed in globules on the leaves and grape berries and dried in large drops. This condition was reflected in the percentages of infested grapes at harvest time, 17 per cent on Concords and 25 per cent on Catawbas. This increase in infestation on the Catawba variety did not occur with the laundry or resin soap, and so seems to indicate less adhesive power late in the season in the soft soap, since the Catawbas were harvested 12 days later than the Concords. The laundry soap used (PL IV, fig. 1) was the com- mon yellow-bar soap, chipped and dissolved in hot water. This spread smoothly over the grape foliage and berries and gave a satis- factory covering when the spray was directed on the clusters for a sufficiently long time, but the amount required to " wet " them was 33 per cent greater than when the resin soap was used. Wlien the third-spray application is used, as in this experiment, the adhesive I CONTROL OF GRAPE-BERRY MOTH. 21 quality of the laundry soap appears equally as great as that of the resin. However, when the first and second sprays are applied and the third omitted, leaving a longer period between the last spray application and harvest time, it appears from field observations that the resin soap adheres longer than the laundry soap. The resin fish-oil soap used was the commercial product obtained in bulk and of the consistency of thick molasses. This soap is readily dissolved in hot water and wets the clusters (PI. IV, fig. 2) easily, as is indicated by the use of but 75 gallons as compared with 100 gallons of laundry soap solution and 140 of soft soap solution. The resin soap adhered the longest of any material tried. It was found that 1 pound of tliis soap to 50 gallons was about as efficient as 2 pounds of the other soaps and at the rate of 1 pound to 50 gallons is as cheap. No difference in compatibility with Bordeaux and ar- senate of lead could be noted among the different soaps. In conclusion it can be said that the resin fish-oil soap proved to have all the desired qualities of a spreader and adhesive and in the present state of knowledge appears the best spreader to use in grape spraying.^ Combination Sprays. In the Sandusky and Lake Erie island sections where the Catawba variety predominates it is desirable to combine a fungicide for con- trol of downy mildew, Plasmophora viticola, with the arsenical and soap for rootworm beetle and grape-berry moth control. In the other sections it is sometimes desirable to use the same combination for blackrot and insect control. Bordeaux, either 2-2-50, 2-3-50, or 3-3-50, was used in combination with arsenate of lead powder IJ and 2J pounds to 50 with soaps at the rate of 1 and 2 pounds to 50. In some of the experiments the copper sulphate was omitted and stone lime, 2 pounds to 50 gallons, was used. The combining of the insecticide with the fungicide appeared to make no difference in insect control. In some cases slight burning of Concord and Catawba foliage and serious burning of Ives foliage resulted from application of the Bor- deaux-arsenate of lead-soap combination. The burning was most noticeable during the abnormally wet season of 1917. Experiments were conducted in 1918 to determine the material or combination causing the burning. The combinations of arsenate of lead and soap with Bordeaux proved responsible. Wherever the copper sulphate was omitted and the arsenate-soap-lime mixture was used, no injury resulted. 'An appreciable difference was noticed in the length of time required to " wet " the clusters of diffei-ent varieties. Beginning with Niagaras, wliich were most readily wet, the other varieties followed in about this order : Catawbas, Delawares, Ives, and Concords. This difference is apparently closely correlated with the waxy bloom on the grape berries. 22 BULLETIN 837, U. S. DEPARTMENT OF AGRICULTURE. The burning was closely related to the thrift of vinesj the stage of grape growth when sprays were applied, the weather during and following spray aplications, and the method of mixing materials. Weak vines and those bearing too heavy crops were most seriously burned. Spray applications just before and after bloom caused more injury than later applications. Excessively wet and cloudy weather during and following spray applications appeared to in- crease burning. When either of the ingredients of Bordeaux was added to the other without being diluted, increased burning resulted. From the above observations it is concluded that Bordeaux mixture should not be used in the arsenate of lead-soap combination on the Ives variety at any time and that in applying the combination with Bordeaux to Concords and Catawbas the above factors influencing foliage injury should be kept in mind. The arsenate of lead-soap- lime mixture was safe wherever used, even on the Ives variety. The injury from spray materials appears to be cumulative from season to season. The combinations of spray materials and factors influ- encing grape foliage injury warrant further experimentation. Dusting for Control of Grape-Berry Moth. Much interest is centering in the application of insecticides and fungicides in dust form as compared with the liquid application. In an attempt to avoid all spray residue on grapes at harvest time, grape-dusting experiments were conducted in 1916, 1917, and 1918. The final infestation in check plats adjacent to the dusted plats was so light as to give inconclusive results except in 1916. The plan of the 1916 experiments and the results recorded are presented in Table IX. Table IX. — Dusting experiment for control of grape-'berry moth, Schonhardt Vineyard, Venice, Ohio, 1916. Dust and spray mixtures used and dilutions. Spray applica- tions. Counts of infested grapes at harvest, Oct. 13, 1916. Plat No. 3 to 5 days after grapes bloom, June 29. When second- brood larvae begin to hatch, Aug. 4. Num- ber of vines exam- ined. Num- ber of clus- ters exam- ined. Num- ber of grapes exam- ined. Num- ber of clus- ters in- fested. Num- ber of grapes infest- ed. Per- cent- age Of grapes in- fested. 1 Arsenate of lead powder, 2J pounds to 50 gallons; Bor- deaux, 3-3-50; laundry soap, 2 pounds to 50 gallons. Li- X Dust mixture, arsenate of lead powder 10 per cent, hy- drated lime 90 per cent X x 8 6 8 2n 215 239 6,330 6,450 7,170 211 202 239 2,978 1,194 5,668 47.04 2 3 Arsenate of lead powder, 24 pounds to 50 gallons; Bor- deaux, 3-3-50; laundry soap, 2 pounds to 50 gallons X 18.51 79.05 CONTROL OP GRAPE-BERRY MOTH. 23 The dust materials were applied with a small hand duster and the liquids with a gasoline power sprayer. In each case the appli- cation was made from either side of each row and a thorough cover- ing of foliage and fruit effected. The dust materials adhered to the grape foliage fairly satisfactorily but did not adhere well to the smooth surface of the grape berries. No rain fell from the time of the application until August 11 when a light shower occurred. When the vines were examined on August 19 only a trace of the dust material was in evidence on the foliage or fruit, while the sprayed fruit was well covered with spray material. It required 40 pounds of dust material to dust 46 thrifty Catawba vines. At this rate and with the vines set 900 to the acre as is the practice in this section, it would require T83 pounds of material per acre. No doubt this would be materially reduced if a power machine were used for the dusting. If but half as much material were required per acre the amount of arsenical would be from 6 to 7 times as great as when applied in liquid form at the rate of 1^ pounds of arsenate of lead powder to 50 gallons of spray and the liquid applied at the rate of 200 gallons per acre. The writers feel that the dust would have to be applied much more frequently than the liquid to be ef- fective for berry moth control. This method of application might be satisfactory for treating small home grape arbors when applied frequently. Spray Residue on Grapes at Hakvest Time. Throughout these investigations records were kept on the com- parative amounts of spray residue on the grapes at harvest time. In all cases where the spray application shown as the third (fig. 1) was used in early August the fruit was heavily coated with spray material at harvest time. In nearly all cases where the combination of first and second sprays was used, and spraying completed by July 25, there was not sufficient spray residue at harvest time to affect the marketing of the grapes in baskets for table use. When either the first or second application was used alone the residue was lighter than when both were used. Slightly more residue resulted on the plats sprayed with arsenate of lead at the rate of 2| pounds than on those sprayed with the same material at the rate of 1^ pounds to 50 gallons. No difference could be seen between the fruit from plats on which Bordeaux was included and those on which lime, 2 pounds to 50 gallons, was substituted for it. Slightly less residue was present on the plats sprayed with arsenate of calcium than on those sprayed with arsenate of lead where the comparison of material was on the basis of arsenical content. In one plat where these materials were mixed in the proportion of arsenate of calcium 9 ounces to arsenate of lead 5 ounces, the amount of residue was 24 BULLETIN 837, U. S. DEPARTMENT OF AGEICULTITRE. greater than where arsenate of calcium was used alone and less than where arsenate of lead was used alone. Summary of Results with Spray Materials. Arsenicals. — Arsenate of lead powder at the rate of 1| pounds to 50 gallons proved adequate for commercial control of the grape-berry moth in the average case. Arsenate of calcium proved almost equally as efficient as arsenate of lead when compared on the basis of arsenical content and has the additional advantage of leaving less residue at harvest time. Spreaders and adhesives. — Resin fish-oil soap at the rate of 1 pound to 50 gallons possessed all the qualities desired and required 33 per cent less spray material than laundry soap and 86 per cent less than soft soap, to wet the grape clusters on an equal area of vineyard. SjJray comhinations. — The mixture of arsenate of lead and soap with Bordeaux should be used with care on Catawba and Concord varieties. The Bordeaux mixture should be omitted on the Ives va- riety. Stone lime at the rate of 2 pounds to 50 gallons should be added to the arsenate of lead-soap combination when Bordeaux mixture is omitted. Dust TTiixtures. — ^The dust mixture of arsenate of lead and hy- drated lime did not adhere to the grape clusters as well as the liquid sprays. The dust material was only partially effective for the con- trol of the grape-berry moth. Spray residues. — Objectionable residues do not result when the first and second spray applications are used with care. A spray ap- plication in August with the materials necessary for berry-moth control will leave a residue which will bar the fruit from the basket market. COST OF TRAILER SPRAYING. Because of the fan training system it was necessary, when spray- ing, to drive between each two rows of grapes. Each rodman sprayed but one side of one row at a time. In the Chautauqua-Erie belt it was found possible for a man to spray both sides of a row as he went, but there appears to be little gain in time by the latter method. In all of the experimental work it was found possible to mix and apply 6 tank loads of 150 gallons each or a total of 900 gallons in 9 work- ing hours. This amount of material covered from 3 to 8 acres, de- pending on local conditions, and averaged about 5 acres. About one- half more material was required for the second application than for the first. Wliere the third application was made on plats that had received the second, the amount was about the same as for the CONTROL OF GRAPE-BERRY MOTH. 25 first application, but where the second had not been applied slightly- more material was required for this third application than for the second application in adjoining plats. This difference is accounted for by the fact that the material from the second application re- mains on the grapes and overcomes the waxy bloom, thereby allowing quick wetting. The following comparisons of single-nozzle, double-nozzle, and spray-guns for use in trailer spraying were made : Table X. — Experiments with single and double nozsles and si;ray guns for use in trailer spraying of grapes. E. Dimning's vineyard, Avon Lake, Ohio, 1918. [First spray application for grape-berry moth control, June 19.] Plat No. Number nozzles per rod. Nozzle aper- tures. Pressure per square inch. Time to spray 150 gallons. Number rows sprayed. Percent- age of time saved. Percent- age of material saved. 1 1 2 Inches. Pounds. 175 175 200 Minutes. 64 50 38 14 16 16 0 21.8 40.6 0 14.2 14.2 2 3 I Spray guns, 1 to each hose line. In Table X it is seen that two medium nozzles per rod saved 21 per cent in tinie and 14 per cent in materials as compared with one large nozzle. SjDray guns saved 40 per cent in time and used no more material than two nozzles per rod, but an angle at the nozzle end of a rod is a necessity for thorough covering of the grape clusters. The writers believe that for the average vineyard two disk nozzles, at an angle of 45°, to each rod, with ^^g-inch apertures and a pressure of 175 pounds, will be found most satisfactory. Materials and labor vary so greatly from season to season and in local sections that figures as to the cost of spraying are of little value. The statement can be made that an average of about 5 acres of thrifty vineyards can be sprayed by two men with a team in a day and will require from 100 to 250 gallons, averaging 147 gallons (Tables I and II) per acre for the first application, and from 160 to 300, with an average of 224 gallons per acre, for the second applica- tion. CONCLUSIONS. The grape-berry moth has been a more general pest in northern Ohio than in other commercial grape sections because of the follow- ing conditions: (1) Production of the late maturing Catawba va- riety, (2) cultural methods favorable to successful wintering of the insect, (3) harvesting methods which leave the insect in the vine- yards, (4) a grape training system which prevents spray materials from reaching the clusters when applied with set nozzles. 26 BULLETIN 837, U. S, DEPARTMENT OF AGRICULTURE. Spray schedule. — The combination of first and second spray appli- cations is adequate for control on the principal varieties of grapes grown in northern Ohio and when carefully applied leaves the fruit suitable for the basket market. Spray ■materials. — ^A combination of arsenate of lead powder 1| pounds to 50 gallons and resin fish-oil soap 1 pound to 50 gallons, in Bordeaux mixture or with stone lime 2 pounds to 50 gallons, may be used for spraying Concords and Catawbas. Copper sulphate should not be used in the above mixture for Ives variety. Arsenate of cal- cium, commercial powder, proved almost as efficient as arsenate of lead for grape-berry moth control. Dust mixtures do not adhere to the grape berries as well as liquid sprays but may be used on small home grape arbors if applied frequently. Spray residues. — The grapes will be practically free from spray residue if the schedule recommended is used according to directions. Sprmj mctlwd. — The trailer method only was used, and a trailer provided with a short rod and two angle nozzles proved most satis- factory in most vineyards. RECOMMENDATIONS. When possible, vineyards should be placed in condition for winter at the end of the cultivation season in July and left without further cultivation until spring ; this practice is designed to increase the win- ter mortality of the grape-berry moth pupae. Nwmber of spray applications. — For general practice for grape- berry moth control in northern Ohio two spray applications should be made. Time. — The first application should begin 3 to 5 days after grapes set and the second should begin when the grapes touch in the clusters. This second application will usually come 3 to 4 weeks after the first. Method. — Where the berry moth is a major pest the trailer method of spraying is the only one that will give complete control. Materials. — Arsenate of lead, at the rate of 1^ pounds of powder or 3 pounds of paste to 50 gallons, as the active killing agent, with resin fish-oil soap, at the rate of 1 pound to 50 gallons, for a spreader and adhesive, used either in Bordeaux mixture or with 2 pounds of freshly slaked lime to each 50 gallons, has proved the most consistent combination tried. Bordeaux mixture should not be used on the Ives variety of gi'apes in northern Ohio because of the danger of injury to the foliage. Amounts of material should be great enough to allow the covering of all clusters with a thin, smooth film of spray material. UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 838 Contribution from tlie Bureau of Entomoiogy L. O. HOWARD, Clilef Washington, D. C. PROFESSIONAL PAPER June 5. 1920 CYPRESS BARK SCALE.^ By F. B. Herbebt, Scientific Assistant, Forest Insect Investigations. CONTENTS. Economic importance History Native host plant Distribution and spread- Injury Associated insects Food plants Page. 1 2 2 3 4 7 Descriptions Life history and habits Seasonal history Predacious and parasitic enemies- Control experiments Recommendations for control Summary Page. 8 12 16 18 19 20 22 ECONOMIC IMPORTANCE. The Monterey cypress (Cupressus macrocarpa Hartw.) is one of the most popular shade and ornamental trees in California. It is planted separately or in hedgerows and often trimmed to formal shapes. Due to its thick, spreading habits it makes a good wind- break where it is planted in exposed areas. It is used especially along the coast and in the sandy citrus areas of San Bernardino County. It is also cultivated as an ornamental tree in many other parts of the world. The cypress bark scale infests a large percentage of these trees in California, causing a great deal of injury, particularly to the thickly planted hedgerows (PI. I, figs. 1, 2) and windbreaks. In the San Francisco Bay region it ranks first among the pests of the ^ Ehrhomia cupressi (Ehrhorn). Order Hemiptera, suborder Homoptera, family Coc- cidae. Note. — An investigation of the injury to cypress trees in California was taken up by the writer, at the suggestion of Dr. A. D. Hopkins, Forest Entomologist, in November, 1916. An examination made of these trees in the vicinity of the Los Gatos laboratory disclosed the main cause of the injury to be the cypress bark scale, Ehrhornia cupressi (Ehrhorn). This bulletin contains a record of studies of its history, biology, importance, and control made during this investigation. The writer wishes to acknowledge the assistance of Dr. A. D. Hopkins, Forest Ento- mologist ; Mr. H. E. Burke, Specialist in Forest Entomology ; Prof. R. W. Doane and Mr. G. F. Ferris, of the Stanford University Department of Entomology ; and Mr. George P. Gray, of the University of California Insecticide Laboratory, all of whom offered helpful suggestions during this study. Mr. R. D. Hartman assisted in the control experiments. 150829°— Bull. 838—20 1 2 BULLETIN 838, V-. S. DEPAETMENT OF AGRICULTURE. Monterey cypress. The insect is extremely difficult to control — a fact which makes it a very disagreeable and harmful pest. HISTORY. The cypress bark scale was first described in 1911 as Sp/uiero- coccus cupressi by Mr. E. M. Ehrhorn ^ who collected it at Niles and San Jose, Calif., as early as 1903, and at Belvedere, Calif., in 1908, in the bark crevices of Monterey cypress. The next mention of this insect was in the report of the Selby Smelter Commission in 1915, when Mr. J. W. Blankinship ^ and Prof. R. W. Doane ^ reported it as one of the factors causing the death and dilapidation of the Mon- terey cypress in the Selby smoke zone. In October, 1918, Mr. G. F. Ferris,^ with the writer's approval, erected a new genus, Ehrhornia, for this species, for it certainly was not a Sphaerococcus. Although an important shade-tree pest, no discussion of this insect has ap- peared in print otherwise, except for a short note by the writer* on its damage and distribution. This coccid has no synonyms, having been listed under the name /Sphaerococcus cupressi until the new genus, Ehrhornia, was erected in 1918, with cupressi as the type. NATIVE HOST PLANT. This scale insect could not be found at Cypress Point or Point Lobos, Calif., the only localities where the Monterey cypress is native, which showed quite conclusively that this tree was not the native host. On November 9, 1917, the cypress bark scale was found infesting some planted trees of Arizona cypress {Cupressus arizonica) at San Jose, Calif., although it has not at the present writing been found on the native cypress in Arizona. On December G, 1917, this insect was found by the writer infesting one incense cedar {Libocedrus decurre^is), and later several more trees, on the Stanford University campus. Later, the native incense cedars at Placerville, Calif., were examined. A number of them were found infested. They were several miles from any planted cypresses, which, moreover, were not infested. Later, the cypress bark scale was found upon incense cedar at Ashland, Oreg., by Mr. Albert Wagner, of the United States Bureau of Entomology; at Crockers, Calif., by Mr. G. F. Ferris; and at ^ EhrhorNj E. M. new coccid.e with notes on other species. In Can. Ent.. v. 43, no. 8, p. 275-280. (Figs. 3, 3a, b, c.) 1911. 2 Holmes, J. A., Franklin, E. C, Gould, R. A. report of the selby smelter com- mission. U. S. Dept. Int., Bur. of Mines, Bill. 98, p. 381-397 ; 428-450. 1915. 3 Ferris, G. F. notes on coccid.e II. In Can. Ent. v. 50, no. 10, p. 32.3-332. 1916. ■* Monthly Letter of the Bureau of Entomology, U. S. Dept. Agr,, no. 46, p. 5. February, 1918. I CYPRESS BARK SCALE. Stirling City and Giant Forest, Calif., by the writer. These localities are all from 75 to 150 miles apart and from 5 to 50 miles distant from any planted cypress. This seems to indicate that the incense cedar is the native host for this scale insect. Mr. G. F. Ferris reports finding this scale insect upon herbarium specimens of Guadalupe cypress {C. guad-alupensis) in the Stanford University herbarium, collected on Guadalupe Island, Mexico. It is impossible at present to state whether this is one of the native hosts of the scale, or whether the scale has been carried there from the mainland. DISTRIBUTION AND SPREAD. Incense cedar, the original host plant of the cypress bark scale, occurs in California, Oregon, western Nevada, and Lower California. The majority of these trees are found in the Sierra Nevada and northern coast range mountains of California, and in the Siskiyou, southern coast range, and Cascade Mountains of Oregon. The cypress bark scale has been found at four widely separated points in the Sierra Nevadas (Stirling City, Placerville, Crockers, and Giant Forest, Calif.) and at one point in the Siskiyou Mountains (Ashland, Oreg.), which indicates that the scale undoubtedly occurs throughout the major portion of the range of the original host. From this host the scale insect has spread to planted incense cedars and cypresses in the more thickly populated regions of the State. Two probable methods of distribution are suggested. The cypress bark scale has been found to thrive on very young trees. Incense cedar seedlings occasionally are brought down from the Sierra Nevadas by tourists to plant in their own jards, and it is quite possible that the scale was carried to the valley on some of these trees. Rustic incense cedar is also transported from the Sierra Ne- vadas to be used quite extensively for pergolas and porch pillars- Orclinarily this would be done during the summer, which is the re- productive period for the scale insect. As the females can live for some time on green logs, it would be very easy for the young larvae, hatched en route or after the logs have reached their destination, to attach themselves to near-by cypress trees, and thus start a heavy infestation in a new location. From these original points of infestation the pest has spread through large areas. This has been accomplished for short distances by the usual agencies of wind, birds, insects, etc., and for longer dis- tances by the shipment of infested nursery stock. The insect has been found by the writer infesting cypress seedlings in nurseries. Close planting of shade trees, hedges, and windbreaks undoubtedly aids in its rapid spread by natural means. 4 BULLETIN 838, U. S. DEPARTMENT OF AGRICULTURE. The scale insect, besides being distributed on the incense cedar, has now spread until it occurs on a large percentage of the cypress trees and hedges in almost every locality about San Francisco Bay, par- ticularly on the San Francisco Peninsula, in the Santa Clara and Livermore Valleys, and north of the bay in Solano, Marin, and So- noma Counties. It has been found in one locality. Riverside, in south- em California, where a great many trees are planted for windbreaks. It is also to be found on Guadalupe Island, Mexico, as it was recently taken from herbarium specimens of Guadalupe cypress collected on this island. The accompanying map (fig. 1) indicates the localities in which the cypress bark scale has been found to date, as well as the range of incense cedar and Monterey cypress. There are many localities within the range of cypress and incense cedar, which the writer has not yet visited, in which the scale insect will probably be found, when investigated. In all probability it eventually will infest all planted cypresses unless radical measures of control are adopted. Since it has been found that the cypress bark scale can live on Arizona cypress, it is possible that it may spread to that host in Arizona and Mexico, or it may even be able to adapt itself to closely related hosts and spread throughout the country. INJURY. Injury to the tree is caused by the myriads of insects which are to be found in every crack and crevice of the trunk, branches, and twigs, each sucking out the plant juice through its long thread-like mouth parts. Under each scale may be found a small brown ring in the cambium, showing the tissue killed by each individual. There is no secretion of honey dew, except for a small amount by the young larvae, and only a slight formation of black sooty fungus about these insects, but a secretion of white cottony wax pro- truding from the bark crevices and covering the twigs gives abundant evidence of their presence (PI. II). First a limb or two on an infested cypress turns yellow, then red or brown, giving the tree (PI. Ill) a scraggy appearance. This appearance often starts near the top of the tree and works down toward the center, or perhaps spreads from one limb to the rest until the whole tree is dead. Quite often the trees are dug up or felled before this final stage is reached; others are left to mar the land- scape until they rot and fall. In hedges, yellow and red spots appear, which increase finally to large proportions, leaving wide gaps of dead material which eventually destroy the beauty of whole hedges. One hedge in Livermore, nearly a half mile long, was infested and dead or dying CYPRESS BAPJv SCALE. Mar 0? DISTglBOTIOH. OBISISVL liAlrCZ 0? IKCEHSE CEDAR ■^»-.if-\'\.l OEIOIKAL AMD PIAHTED EaHGE OF MOHTEEKT CYPRESS lOCUItISS EHOWH TO BE INFESTED WIIE C^KE33 BAHK SOAIE. s:^ Fig. 1. — Map of distribution of the cypress barli scale. 6 BULLETIN 838, U. S. DEPARTMENT OF AGRICULTURE. for its full length. Young hedges but 3 or 4 years old have been found infested and some of the trees dead even before the hedge has become of any particular use or ornament. Windbreaks with an opening here and there, caused by the death of a tree or group of trees, are not nearly as efficient as they would be otherwise. In the drier regions of the State the injury is more apparent than in the fog belt, where the tree seems to be much more thrifty. In tine former regions the cypress is not a long-lived tree, and when in- fested death is hastened considerably. Trees occasionally are found that are heavily infested, yet quite normal in appearance. It is believed that such trees have only re- cently beeii infested, and will eventually show the effects of this slow- working insect. In some localities in central California there is hardly a respect- able row of trees or hedge left to greet the eye. At Los Gatos many trees have died and very few remain which are not now infested. About San Jose, Livermore, Benicia, etc., there are also heavy in- festations. INJURY IN THE SELBT SMOKE ZONE. The " Selby smoke zone '^ is an area extending for nearly 10 miles along the Carquinez Strait, between San Pablo and Suisun Bays. The Selby smelter is located on the south side and at the west end of this strait. The prevailing w^inds are from the west and southwest, thus blowing the smelter smoke across the strait onto the territory between Vallejo and Benicia. In this area there has been considerable complaint of damage done to different trees and plants by this smoke. Many of the Mon- terey cypresses in this territory are dead or dying. Examination of these trees and plants by specialists of the Selby Smelter Commis- sion has proved that insects and fungi are responsible for part, at least, of the damage. The writer has examined Monterey cypresses in certain parts of the smoke zone and has found the scale insect abundant. In the Benicia Cemetery, practically 100 per cent of the trees were found infested and a large percentage were dead or dying. According to Prof. Doane, this cemetery, although infested in 1913, contained but few dying trees. Probably at that time the infestation was rather re- cent, but has gained headway since. In the writer's judgment, the cypress bark scale is the main factor causing the unsightl}'' and dying condition of these trees. One need only see the condition of the cypresses in the Livermore and Santa Clara Valleys to reach this conclusion. CYPRESS BARK SCALE. 7 ASSOCIATED INSECTS. The cypress bark scale is by no means the only enemy of the cypress. The cypress barkbeetles {Phloeosirms cu-pressi Hopk. and P. cristatus Lee.) are important primary insects, causing the death of a considerable number of trees, and perhaps ranking first among the pests of the cypress in California, considering the State as a whole. In the San Francisco Baj^ region they must, however, take a place second to the scale in point of damage done. Many people throughout the State have considered the beetles as the only enemies of the cypress, and, when noting the death of a tree, have taken it for granted that the beetles were the primary cause, without bothering to investigate properly. The entire foliage of trees killed by the beetles turns first yellow, then brown, and is much more conspicuous than the foliage of a tree being killed by inches by the scale insect. The presence of the beetles is also more easily detected. At times the beetles have been found working independently of other insects, and killing trees. For example, at San Carlos, on the San Francisco Peninsula, the beetles have been killing several trees per year for a number of years. In the spring of 1918 they were found entering the green trunks of live trees and girdling them. Other trees showing the work of Phloeosinus, dead one or two years, stood near by. The recently attacked and the unattacked trees were apparently very healthy. The beetles are often secondary pests, entering trees well infested and weakened by the scale insect. At Martinez, Calif., in January, 1918, a row of 12 trees was heavily infested with the scale insect. Three of these had been recently killed by barkbeetles. On Alum Rock Avenue, San Jose, Calif., there is a long double row of cypresses, all of which are heavily infested with the cypress bark scale and practically all of which have a sickly appearance. Here the barkbeetles attack an occasional tree, or a section thereof, and hasten its death. They also impair the beauty of the trees by entering into the center of small twigs and weakening them, so that the wind breaks them off. When gathered together, the twigs from one tree formed a pile 2| feet high and nearly as broad. There were also many twigs still hanging broken in the tree. The injury to cypresses by these beetles will be treated separately in a later paper. Three species of mealybugs, Pseudococeus ryani Coq., P. sequoiae (Coleman), and P. curpressicolus Ferris, also infest cypresses and occasionally do some damage. Other associated scale insects are Xyl^coccus moGWcarpae Cole- man, Lecanium corni Bouche, Diaspis caruell Targ., Aspidlotus 8 BULLETIN 838, U. S. DEPARTMENT OF AGRICULTURE. hederae (Vallot), and Aspidiotus ehrhomi Coleman. None of these have been noted doing any considerable damage. Still other associated insects of various orders are: Phymatodes nitidus Lee, Atimia confusa Say, Trachyhele hlondeli Mars., the cypress moth {Argyresthia cwpressella Wals.), the cypress cone- borer {Cydm cupressana Kear.), a horn-tail wasp {Sirex calif omi- cms Ashm,), the arborvitee plant-louse {LacJiniella tujaflina Del Guer.), and an undetermined tussock moth. FOOD PLANTS. The known food plants of the cypress bark scale are: Monterey cypress {Cwpressus raacrocarpa Hartw.), Arizona cypress {C. ari- zonica Greene), Guadalupe cypress {C. guadalupensis Wats.), and incense cedar {Libocedrus decurrens Torr.). On one other tree, a deodar cedar {Cedrus deodara Loud.), at Santa Eosa, Calif., a dead male was found in its cocoon. It seems strange that the scale insect should not occur on all species of cypress if it will infest two trees as different as Monterey cypress and incense cedar, yet Italian and Oriental cypresses, two varieties of Cupressus semperoirens^ are immune to the attack of this insect. They have been found in many instances in close proximity to infested Monterey cypresses and entirely free from the scale insect. In one case, in the Benicia Cemetery, 27 cypresses formed a square about a plot. Two-thirds of these were Monterey cypresses, with every third tree an Italian cypress, touching a Monterey cypress on each side. Every Monterey cypress was infested and dead or dying, while not a scale could be found on the Italian cypresses. Specimens of Himalayan cypress {C. torulosa Don.), Macnab cy- press {C. mcwTiabiana Murray), funeral cypress {C. fwnebris Endl.), Sargent cypress {C. sargentii Jepson), and Port Or ford cedar {Chamaecyparis lawsoniana (Murr.) Pari.) have been examined, although not in large numbers, within the infested areas, and no cypress bark scales could be found upon them. DESCRIPTION.! THE EGG. Egg (PI. IV, A), immediately after being deposited, regularly oval, smooth, and shiny, of a transparent pale yellow color, with eyes of emi)ryo visible through membrane as two dark spots near one end. Average length of seven eggs 0.34 ; mm. ; width 0.14. LARVA. FIRST INSTAR. Young larvae (PI. IV, B) of both sexes alike. Pale yellow in color, with long, flat, oval bodies 0.43 mm. in length, 0.20 mm. in width. Antennte (PI. V, 1 The following detailed description of all stages from the egg to adult, both male and female, were made from living and freshly mounted material collected during the study of the cypress bark scale. Bu!. 838, U. S. Dept. of Agriculture. Plate n r' .^1 % .K, •''■ *'V4^j Fig. 1— a Tall Monterey Cypress Hedge, Showing Many Dead Spots Caused by the Cypress Bark Scale. ^■^ ^^f| Fig. 2.— a Monterey Cypress Hedge Killed by the Cypress Bark Scale. THE CYPRESS BARK SCALE. Bui. 838, U. S. Dept. of Agriculture. Plate 1 1. Fig. I.— Characteristic Infestation on Monterey Cypress Twig, Showing THE Cottony Excretions of the Cypress Bark Scale, x 0.4. Fig. 2.— Characteristic Infestation on the Bark, the Cottony Excretions OF the Cypress Bark Scale Protruding from the Crevices, x 2. ^"- W^:>^>^ *t?^^: ; Sfc«».Nfc A-^' Fig. 3.-THE Dry Outer Portions of the Bark Removed, Revealing the Females of the Cypress Bark Scale Underneath, x 4. Bui. 838, U. S. Dept. of Agriculture. Plate III, Fig. I.— a Monterey Cypress Which Is Nearly Dead from the Work of THE Cypress Bark Scale. Fig. 2.— a Row of Monterey Cypresses, All of Which Are Practically Dead from a Cypress Bark Scale Infestation. THE CYPRESS BARK SCALE. Bui. 838, U. S. Dept. of Agriculture. Plate IV. The Cypress Bark Scale. ^,Egg(X55); £, larva, first instar (X 85); C, female larva, second instar (X55); D, adult female (X 55); E, portion of bark removed, showing adult female depositing eggs, the eggs hatching, and the larva crawling away. CYPRESS BAEK SCALE. 9 A, 2) six-segmented, first segment broadest, sixth longest. Normal antennal formula as follows: 6, 1, (2, 4, 3, 5). Joints 2, 3, 4, 5 subequal and variable, thus causing various formulte. Antennae rather hairy. Fifth segment bearing one broad spine, while four occur on sixth segment. Eyes marginal, set a short distance behind antennce. Legs rather short and stout. Tarsus slightly longer than tibia, the t\^'0 combined slightly longer than femur and trochanter combined. Legs rather hairy, tarsi bearing four knobbed digitules. Usual coccid sucking mouthparts present. About 26 trilocular pores set rather regularly about margin and two longitudinal rows of the same type on dorsum of insect. Caudal setse (PI. V. B, 1) prominent, rather stout, and about 140 fi long; borne on small inconspicuous anal lobes. Anal ring bearing six setse, noticeably longer than in later stages. A number of short spines scattered over body. SECOND INSTAR. Immediately after the first molt some differences between the sexes can be detected imder the microscope. Toward the end of the second stage an ex- ternal difference can be observed in the shape, the female being broader than the male. Female larva (PI. IV, C) about 0.84 mm. long and 0..51 mm. wide, oval in out- line, yellowish brown in color. Legs and eyes similar to those of first-stage larva. Antennae very similar except that sixth segment bears only two broad spines. Caudal setae (PI. V, 5, 3) much shorter, being only 37 ;oi long. Anal ring still bearing six hairs, also shorter than in preceding stage. A larger number of tubular wax ducts, varying from 40 to 7.5, to be found on dorsum, particularly abdomen, and on lateral margin. Also a number of trilocular pores and small spines scattered over body. Male larva (PI. VI, A) about the same length but narrower than female; 0.85 mm. long by 0.43 mm. wide. Color yellowish. Caudal setae twice as long as those of female, half as long as those of first-instar larva, 75fi. Six anal hairs, in length equaling diameter of anal ring; longer than those of second-stage female, but shorter than those of first-stage larva. Pores of male larva very inconspicuous (PI. V. D, 2), smaller than those of female. Small quinquelocu- lar type pores scattered over both dorsum and venter of body. Small tubular ducts found mostly on margin. Small spines also present. ADULT FEMALE, THIED IN STAR. Body (PI. IV, D) nearly circular and quite convex, the width exceeding the depth and the length exceeding tlie width. Average length 1.4.5 mm., average width 1.35 mm. General color reddish brown. Anterior half of body quite heavily chitinized, particularly on margin. Derm smooth. Antennae (PI. V, A, 3) no longer marginal, but occurring on ventral side of body, six-segmented, slightly longer, averaging 144 fi, but similar to those of second-instar larva, bearing only two of the four spines found on sixth segment of first-instar larva. As in larva antennal formula varying considerably, of practically no value. Average formula as follows: 6, 1, 3, (5, 2, 4). Eyes lacking. Legs (PI. V, C, 1) short and stout. Tibia and tarsus (PI. V, C, 2) subequal, being together slightly shorter than trochanter and femur combined. Leg bear- ing a few hairs and tarsus bearing four knobbed digitules. Usual coccid mouth- parts present. Large tubular ducts (PI. V, D, 1) occurring on margin and dorsum of body. Small trilocular pores and fine spines scattered over body, particularly on abdomen. Anal ring (PI, V, B, 4) occurring on ventral side of body, small, simple, and with six small setae or hairs. A pair of small caudal setae one on each side of anus. Anal lobes absent. 150829°— 20— Bull. 838 2 10 BULLETIN 838, XJ. S. DEPARTMENT OF AGRICULTURE. MALE PREPUPA. With the second molt, the male larva assumes the form of a partly developed pupa, called the prepupa (PI. \I, B). Form elongate oval; length 0.93 mm., width 0.50 mm. Color pale glassy brown, eyes black. Antennae short, reaching to base of anterior legs and now indistinctly 10 segmented. Wing pads very short, curving under body to middle pair of legs. Anterior legs reaching for- ward, covering " face," the two posterior pairs lying against abdomen. Mouth parts wanting. Segmentation of thorax indistinct. Prepupa incapable of move- ment, except in abdominal segments and anterior legs, which may be feebly moved. MALE PUPA. With the third molt, the male becomes a true pupa (PI. YI, C), greatly re- sembling the adult male, except for the lack of anal wax filaments, and the possession of wing pads instead of wings. General color light brown, head and large, conspicuous wing pads paler, legs and antennae glassy white, eyes black. Antennae distinctly 10-segmented and longer, reaching to base of middle pair of legs. Wing pads appressod to sides of body and extending posteriorly to second abdominal segment. Legs capable of some movement, anterior pair extending beyond the head ; middle femora placed transverse to and extending beyond lateral margin of body, rear pair inclining posteriorly. The larval eyes have disappeared, and have been replaced by one dorsal and one ventral pair. Mouth parts replaced by an approximate pair of eyes. Length 0.95 mm. ; width 0.45 mm. ADULT MALE. (PI. VI, D, E.) Measurements of average adult male: Length of body (exclusive of ap- pendages) 0.75 mm.; antennae 0.55 mm.; wax filaments 0.93 mm.; wing ex- panse 2.15 mm. General color light brown with paler brown appendages. Antennae (PI. V. A, 1) 10-segmented, rather hairy, first joint short and broad, second rather long and broad, others more slender. Antenual formula : 3, 4, 5, 6, 7, 8, 9 (10, 2), 1. Legs rather long, slender, and somewhat hairy. Wings transparent white, slightly iridescent and pubescent. A veinlike thick- ening, beginning at base of wing, branching near base, one branch paralleling costal margin, the other extending toward anal distal margin. Club-shaped halteres each bearing a hook, which catches in a pocket on anal margin of wing. Abdomen terminating in a short blunt style. Two long white wax filaments arising one on either side of base of style, and extending posteriorly. Each filament arising from a number of pores which surround base of two long slender setae. Setae enveloped by wax filaments. SUMMARY. As will be noted from the foregoing descriptions, there are three instars (excluding the egg) in the female. These all have legs, which are not used after the larva is attached. All stages are found in crevices or under some covering on the bark and are nearly or com- pletely concealed by the enveloping cottony Avax. Newly hatched first-instar larvae may be found crawling actively over the bark before attachment. There are five stages in the male. The first two stages are found in similar positions, but the second stage after becoming full grown re- CYPRESS BARK SCALE. 11 moves to a dry, secluded spot, where it spins a cocoon in whicli the remaining transformations take phice. Size and shape are of some use in distinguishing the different stages. The female in its second stage is considerably broader than the male in the second stage, while both are larger than in the first stage. The adult female is more circular than in the preceding stage. The antennae of the adult female are slightly longer than the an- tennae of second-stage larvse, while the latter are slightly longer than those of the first-stage larvae. All are ver^^ similar, however, the only distinguishing character being that the first-stage larva possesses two more broad spines on the sixth segment. The three pairs of legs on each individual are alike, nor is there any difference between the larvas and the adult female, except a very slight one in the relative lengths of the femur, tibia, and tarsus. Mr. Ehrhorn's figures indicate a difference in the arrangement of the hairs and digitules, which the writer has been unable to detect. There is a very small, scarcely discernible, tooth or " denticle " on the face of the claw in all these stages. The caudal setse give good characters for the separation of the different stages. On the first-stage larva they are about 140 microns long; on the second-stage male they are about half as long, 75 mic- rons; those of the second-stage female half as long as the latter, 37 microns ; and those of the adult female very much shorter still. The length of the setae on the anal ring also decreases in the same ratio. Simple marginal eyes are present in all the larval stages, but not in the adult female. Mouth parts are present in all stages, except in the male prepupa, pupa, and adult. Four spiracles are present in all the larvae and the adult female, one behind each of the four forward legs. In the larvae they appear as simple tubes, and more as large chitinized circles in the adult female. There are several types of pores found on the derm of the scale insect, which aid in distinguishing the different stages. The small sessile pores are of two tj^pes, viz., "trilocular" (PL V, Z>, 4) and " ciuinquelocular '' (PI. V, Z^, 5). The former are more or less tri- angular and contain three cells or loculi. The quinquelocular pores are nearly circular, but tend to be five-sided, containing ordinarily six loculi, one in the center with five clustered about it. Aside from these there are circular pores communicating with internal ducts. These are short and tubular, bearing at the inner end a cup- shaped depression. All types are presumably capable of secreting wax. The first-stage larva bears only small pores of the trilocular type. These are arranged in a marginal row on each side of the body and in two longitudinal rows on the dorsum. 12 BITLLETIN" 838, V. S. DEPARTMENT OF AGRICULTURE. On the second-stage female larva are to be found the small tri- locular pores scattered over the body, and from 40 to 75 of the large tubular ducts on the margin and dorsum (particularly on the ab- domen). The pores of the second-stage male larva are of two types: The small quinquelocular type which are found all over the body, and tubular ducts similar to those of the second-stage female larva, but much smaller and less conspicuous, found mostly on the margin. The pores and ducts of the adult female are of the trilocular and tubular types, the first scattered over the body and the latter occur- ring on the margin and dorsum of the body, much the same as in the second-stage female larva. Small spines (PI. V, Z>, 3) are present on all these stages. Viewed from above, these spines are likely to have the appearance of cir- cular pores, but can be soon distinguished by altering the focus of the microscope. LIFE HISTORY AND HABITS. In the early winter adult females are found containing a few eggs. The eggs increase in number and by early spring from 30 to over 100 may be found within each female. During this time the females have increased considerably in size and have become quite heavily chitinized on the anterior half of the body. OVIPOSITION. The embryos in the eggs develop within the body of the female until they are about ready to hatch, when they are expelled (PI. IV, E) . The female is well surrounded by a cottony secretion, but when oviposition is begun the tip of the abdomen is drawn in, leaving a space in which the eggs may remain until hatched. After hatching, the larvse are usually able to find an exit between the cotton and the surrounding bark. Each female is capable of laying 50 to 100 or more eggs. The greatest number of eggs that has been found within a female at any one time is 105. The eggs are laid slowly, covering a long period of time. They are laid during the warmer part of the day at intervals of 7 to 70 minutes. After laying a series of 5 to 10 eggs, the female ceases oviposition for a day or so and then resumes it. Most of the embryos in the eggs are deposited tail first, about one- sixth being deposited head first. The head end of the embryo is evidenced by the two black eyes which are visible through the egg membrane. By a series of contractions of the abdomen, the egg is forced out until entirely free from the body of the adult. These eggs are oblong oval when first deposited, but flatten out somewhat be- fore the rupturing of the membrane occurs. Bui. 838, U. S. Dept. of Agriculture. Plate V. r. A \ B D The Cypress Bark Scale. A AntenniB- 1, adult male; 2, first-stage larva; 3, adult female (same as second-stage larva); 'b Tip of abdomen, showing anal rings and caudal seta: 1, first-stage larva; 2, second-stage m'ale- 3, second-stage female; 4, adult female; C: 1, leg of adult female; 2, tip of tarsus and claw; D, Ducts and pores: 1, large pore and tubular duct of adult female; 2, small pore and tubular duct of second-stage male; 3, spins; 4, trilocular pore; 5, quinquelocular pore. Bui. 838, U. S. Dept. of Agriculture. Plate VI, The Cypress Bark Scale. A, Male larva, second inStar (X 50) (male and female first-instar larvae are identical); B, male prepupa (X 50); C, male pupa (X 50); D, male adult (X 50); E, male adult ready to emerge from cocoon. CYPRESS BARK SCALE. 13 About 15 minutes after the eggs are deposited, the embryo starts a series of convulsions and after considerable struggling ruptures the membrane inclosing it. The ruptured membrane is jDushed down over the abdomen and the larva, which is usually on its back, begins waving its legs about in the air. It usually takes this larva 20 to 30 minutes to free itself from the membrane, and after exercising its legs for 30 or 40 minutes it finally gets to its feet and crawls away (PI. IV, E). The larva quite often is stuck to the next expelled Q,gg and may be held out in space for some time, but the struggles of one or the other finally allow the feet of the larva to touch foundation, where it soon makes use of them. MIGRATION. The larvae become very active soon after exclusion and begin search- ing for a suitable spot upon which to locate. The majority of them immediately work down into the bark crevices or under the cottony secretions of the parent females, where they become attached. Some seem more fastidious than others and travel farther in search of newer feeding grounds. Recently hatched larvae when placed upon favorable young trees do not travel far, and usually settle down after investigating two or three crevices in the bark. Larvae placed upon paper were able to travel considerable dis- tances. (See fig. 2.) The average rate of travel for six larvae was 54.25 cm. per hour, which they were able to maintain for several hours. The greatest distance traveled by one larva was 174 cm. in two hours. One larva, after traveling 124.46 cm. in four hours, ap- parently f)ut its last efforts into trying to pierce the paper with its proboscis. Table I gives the time, distance, and rate per hour trav- eled by six cypress bark scale larvae on black paper for the first few hours of their migration. Black paper was used in order to facili- tate following the tiny pale larvae in their wanderings. ^Vhen white paper was used the larvae were soon lost. There seemed to be a slight phototropism in the case of most larvae, the majority of them finally wandering toward the light. Table I. — Record of travel of six- first-instar larvcc of the cypress bark scale on rather smooth black paper. No. Time. Distance. Rate per hour. Hours. Cm. Cm. 1 1.5 109. 22 72.81 2 2.5 143.51 57.40 3 4.0 124.46 31.11 4 2.0 173.99 86.99 5 2.0 142.24 71.12 6 2.0 66.04 33.02 Average. 2.333 126.577 54.255 14 BULKETIN 838, U. S. DEPAETMENT OF AGRICULTURE. LarvPG isolated in vials, immediately after hatching, lived for two and three days. Living for this length of time and traveling at the above rate of speed during only the warmer parts of the daj, lars^se could go considerable distances in search of proper food. In this way larva? are able to migrate from one tree to another in closely planted hedges or windbreaks. During this migratory period, larv?e are also likely to be transported short distances by dropping from high Fig. 2. — Tracings of five first-iustar larv£e of the cypress bark scale during migration. Reduced 5J times. All were started from tbe same center. branches and being carried by the wind, and for longer distances by animate agencies, such as insects, birds, and animals. ATTACHMENT. As soon as a larva finds a suitable crevice or a jjrotected area in the bark, it thrusts its proboscis into the bark tissues, where it remains permanently. Larvae have not been known to remove themselves from this first location to another after once becoming attached. Larvae w.hich have become detached somewhat later in life are able to crawl about feebly, but finally die without being able to attach CYPRESS BARK SCALE. 15 themselves again. Although the legs are retained throughout the full lifetime, they are of no further use to the female larva except to aid in removmg the cast skins when molting. Larvse will attach themselves on twigs as small as one-fourth inch in diameter and on trunks a foot or more in diameter, provided the bark is not too thick to be pierced with their proboscides. A few scale insects have been found infesting the smooth trunk of seedlings less than one-half inch in diameter, but rough bark is essential to a heavy- infestation. The deeper the larv?e are able to go into the crevices the more satisfied they appear to be. They have been found so well secreted in crevices that it would seem there was no room left for their future growth, and much less any chance of mating, particu- larly after being enveloped with a white flocculent secretion. LARVA. FIEST INSTAR. Immediately after attachment the larva begins enveloping itself with this white cotton until entirely hidden from view. A drop of honcA^dew, resembling pitch, is emitted by some lar^^ae, especially on vigorous trees, during the first few weeks after attachment. Growth starts immediately after attachment and is practically constant throughout the whole instar. The larva at the end of the instar is very similar to those just hatched, except that the former are larger, somewhat broader in proportion to their length, and slightly darker in color. At the end of the first instar, the larva molts, the skin being pushed down off the tip of the abdomen. From 40 to 44 days were required to complete the first instar in the few cases observed. SECOND INSTAE. There is very little development during the second stage. The female larva secretes more waxy cotton and changes in size and shape until it resembles the adult female. After a period of from one to two months the second molt occurs and the larva becomes an im- mature adult. The male larva increases in size and becomes yellowish white in color. It takes on a firmer and trimmer appearance. After a slightly shorter time than that required by the female larva, the male larva detaches itself and crawls about in search of a favorable place in which to pupate. It may pick a spot under some cotton or in a curl of the outer bark. Cocoons have also been found in the cast skins of coccinellicl larvse and in the ruptured bodies of dead female scale insects. 16 BULLETIN 838, IT. S. DEPAKTMENT OF AGRICULTURE. Here the male larva proceeds to spin a cocoon, secreting cottony wax from the small ducts which occur on both the dorsum and venter of the body, turning over and over in the operation. The cotton se- creted is finer than that secreted by the female larva, as would b© expected because of the smaller ducts on the male. It requires three or four days to complete the cocoon. After a day or two of inactivity, the larva molts to a prepupa, pushing the cast skin out through a slit which is in the rear end of the cocoon. MALE PREPUPA AND PUPA. The male from now on is without mouth parts, and during this dormant period is an inactive creature, capable only of feebly waving its front legs and wriggling its abdomen when disturbed. The male remains in this stage from 10 to 15 days, and with this molt becomes a true pupa, greatly resembling the adult male. As in the previous stage, the pupa is inactive. Normally the same length of time is required for this stage as for the previous one. A few pupae have been found hibernating in the colder Sierras. ADULT. When the pupal skin is cast, the male's wings are extended to their full size and then folded, one over the other, upon its back. As soon as the wax filaments have grown to their full length, which requires from 30 minutes to several hours, the male backs out of the cocoon and becomes very active. It immediately begins searching for a mate. The length of life of the male is never more than one or two days and death occurs very soon after mating. FEMALE. The color of the female becomes darker after the second and last molt, and upon becoming heavily chitinized is a dark reddish brown. After mating the body increases considerably in size, becoming nearly globular. Inside the female's body may be found a large number of eggs in different stages of development. The body still is covered quite thoroughly with cotton and deeply hidden in the bark crevices. Upon becoming an adult a new supply of coarser threads of wax is excreted. After depositing the eggs, an act which covers a consider- able period of time, the female shrivels and dies, nothing but the chitinized skin remaining. If the host plant is still alive, this vacancy is soon filled by a female of the next generation. SEASONAL HISTORY. (Fig. 3.). There is but one generation per year, the limits of which are not very definite. The males appear in the fall, being most abundant in CYPRESS BARK SCALE. 17 October and NoA^ember. These mate and die in a few days. At this time most of the females have cast their hist skin and are about one- half grown. The winter is passed as adult females, with no very definite period of hibernation in the lower altitudes. In the Sierra Nevadas there is a more definite period of hibernation and the generations are more even. The female larvae become adults somewhat earlier in the fall. When the cold weather strikes them, development becomes very slow. In December females are found containing a few eggs. These de- velop during the winter and early spring. Oviposition begins on the first warm days of spring and lasts throughout the summer, beginning about April 1 and terminating the latter part of September. In the fall the females, having com- Fig. 3.— Seasonal history diagram of the cypress bark scale. pleted oviposition, shrivel and die. By this time the young females of the next generation are quite well developed, thus assuring the presence of adult females during the whole year. The larvae issue from the eggs 30 or 40 minutes after deposition and soon attach themselves. Larvae of the first instar may be found from April to the middle of October, second-instar larvae from the middle of May to the middle of November, and adult females from about September 15 to the following September. Male prepupse and pupae may be found in September, October, and November, and adults in October, November, and December. A male pupa was found hibernating in the Sierra Nevadas. A few scattering first and second stage larvae may be found during the winter in the milder climate near the coast. 18 BULLETIN' 838, U. S. DEPARTMENT OF AGRICULTURE. Fig. 4. — Nipus Mplagiatus, a coccinellid enemy of the cypress bark scale. X 25. PREDACIOUS AND PARASITIC ENEMIES. There are several coccinellids which aid in the control of the cypress bark scale. None of them is aggressive enough, however, to affect its abundance veiy materially. A very small ladybird, Nipus hiplaglatus Casey (fig. 4), is the most abundant and widespread enemy of the scale. This is a sturdy little beetle, about 1.3 mm. in length, brown- ish black, with a lighter amber spot in the center of each elytron. It is generally present wherever the scale insect is to be found. On one side of a cypress limb in a space 3^ by 24 inches (84 square inches), there were found 46 specimens of this coccinellid and in certain parts of this area there were as many as four beetles to the square inch. There were un- doubtedly still more out of sight under the bark scales. Very few of the larvae of this beetle were seen but this may be accounted for by their extremely small size and their pale brown color. A small black nitidulid beetle, Cyhocephalus califomicus Horn (fig. 5), is very abundant and is sometimes confused with the above coccinellid. It is, however, slightly smaller, shiny black, with more delicate legs, and of a more compact globose form. This beetle has not been seen actually feeding upon the scale insect, yet it is believed to be an aggressive predacious enemy. It is always found about the scale insect, and often with its head in the bark crevices as if feeding upon the scale insect. The larva of this beetle is small and white and not easily found. The twice-stabbed ladybird, Chilocorus hl- vulnerus Muls., is an abundant and aggres- sive predator upon the cypress bark scale. This beetle is often found upon cypress, feed- ing upon this insect pest. The common black-spotted red ladybird, Hippodamia convergent Guerin, has been found a few times feeding upon the scale. Asi this beetle breeds in great numbers in the Sierra Nevaclas, it probably feeds upon the scale on incense cedar in its native haunt. Larvae of the common brown lacewing, Sympherobius angustus Banks, are often found feeding upon the cypress bark scale and aiding materially in retarding its increase and spread. A few specimens of a small hymenopterous parasite have been reared from caged material of this scale insect. It can not, however, Fig, 5. — Cyiocephalus cali- fornicus, a nitidulid bee- tle always found about the cypress bark scale. X 25. CYPRESS BARK SCALE. 19 be considered as of any importance, because of its great scarcity. In all the writer's observations but three scale insects have been found with punctures in them, from which parasites have escaped. Two of them were from Monterey cypress and one from incense cedar in the Sierra Nevadas. The scarcity of this scale insect in the Sierras would indicate that a parasite was quite active upon it. If such is the case, it has not as yet been noted. CONTROL EXPERIMENTS. A series of experiments was undertaken in order to find one or more materials capable of reducing the numbers of this very harm- ful pest. Only those sprays were experimented with which would be able to penetrate well into the crevices of the bark, w^here many of the scales were located. Oil sprays are the only ones which meet these requirements, consequently no others were tried. The oil sprays are capable of penetrating and creeping into all the tiny cracks and crevices of the bark when properly applied. The higher the gi-avity of oil used, the better is the penetration. First, crude-oil emulsion was used, but, being a low-gravity oil, it was unsatisfactory. Next, distillate emulsion was used and, being of somewhat higher gravity, was more satisfactory, but still did not kill more than 40 per cent of the scale insects. Miscible oil No. 1 was next used, being of about 28° Baume gravity. This used in a 12^ per cent solution was quite satisfactory, destroying from 75 to 90 per cent of the scale insects. A well-known washing powder was tried as an emulsifier with both the emulsion and the miscible oil, but proved of no value, so its use was discontinued. Carbolic sheep dip was tried out^ as this was reported to have very good penetrating powers. This, however, gave only about 30 per cent efficiency and was not used further. Six per cent was the highest used with this, as a higher percentage was believed to be dangerous to the tree. It also made the lungs of the experimenter quite sore. Next, miscible oil No. 2 was experimented with, for this had the very high gravity of 33° Baume. A 12^ per cent solution of this proved quite satisfactory, killing a high percentage of the scales and upon second application destroying virtually 100 per cent. Further experiments with this substantiated these results. In Table II are recorded all spraying experiments performed upon the cypress bark scale. All spraying was done in the warm part of the day, generally in the afternoon, when the trees were dry. The spray was applied very heavily and thoroughly, every part sprayed being completely drenched. In experiments Nos. 19, 20, and 21 (small trees) the entire tree was sprayed; in all others only the trunk, lower limbs, and foliage were sprayed, as the trees were too large to be treated with a hand pump. The cypress foliage 20 BULLETIN" 838, TJ. S. DEPARTMENT OF AGRICULTtTRE. appears to be very resistant, for no burning occurred at any time from the application of the spray. It was discovered that the small larvae were much more easily killed than the adult females. All larvse could not be killed with one spraying, however, on account of the long period of hatching. The last of the brood were not hatched by the time the first were becoming adults; consequently two sprayings were necessary, one to exterminate the early hatched larvae, the other to exterminate those hatched later. Table IT. — Re<'ord of spraying experiments performed upon the cypress bark scale. No. Date. Formula. Number of trees. Per cent of efHciency. Remarks. 1 2 1918. Feb. 25 Mar. 9 ...do Mar. 28 Apr. 10 ...do Apr. 24 June 4 July IS Mar. 28 Mar. 29 Apr. 24 ...da...- Oct. 7 ...do July 18 Oct. 7 do Crude-oil emulsion, 7J per cen.t 6 3 5 4 3 3 2 3 2 4 3 2 2 2 2 2 1 1 38 39 1 0 0 0 25 20 20 40 30 30 25 40 75 40 90 75 75 75 100 80 80 85 In shade. Rained that night and fol- lowing 3 days. Do. 3 4 In shade. 5 0 7 g Distillate emulsion, 7 J per cent Distillate emulsion, 7J per cent, and washing powder, 1 pound to 20 gals. Distillate emulsion, 12 per cent Carbolic sheep dip, 3 per cent Warm, dry. Do. Do. Hot. dry. 9 Do. 10 Miscible oil No. 1, 6 per cent Warm, in shade. 11 12 Miscible oil No. 1, 6 per cent,and wash- ing powder, 1 to 20. Warm, dry. Do. 13 14 Miscible oil No. 1, 12i per cent, and washing powder, 1 to 20. Miscible oil No. 1, 12i per cent Do. Warm, dry; repeat on 15 Miscible oil No 1,8 per cent No. 12. Warm, dry; repeat on 16 Miscible oil No. 2, 12i per cent No. 11. Warm, dry; killed 50 per 17 18 do do cent adults and 99 per cent larva?. Warm, dry; rained 2 days ago. Rained 2 days ago; repeat on 1 tree No. 16. Hedge of young trees. 19 Oct. 26 Oct. 28 do do 20 do Do. 21 . ..do Repeat on tree No. 38 of E.xperiment No. 19. In experiments Nos. 1 to 7, percentages are of actual oD content, not emulsion content. RECOMMENDATIONS FOR CONTROL. The following measures are recommended for the control of the scale insect. Cut out all dying trees or limbs of trees beyond saving and destroy them in order to reduce all possible sources of infestation. Purchase clean nursery stock for planting. If the stock is infested, return it to the nursery and demand clean stock to replace it. Most fruit growers now realize that spraying is necessary for the maintenance of healthy trees and the production of clean fruit. Most people, however, still believe that a shade tree should always be able CYPRESS BARK SCALE. 21 to take care of itself. One can not hope to maintain healthful, vigorous, shade and ornamental trees without proper care and occa- sional spraying. Infested trees should be sprayed twice in the fall, once in August or the first part of September and again in the latter part of Sep- tember. This is to kill the larvae before they become mature. The proper dates to spray may vary slightly in different localities and with different seasons, in which case certain phenological events may be relied upon. The first spraying should be done when the fruit of the French prune (the common variety planted throughout the State) becomes blue or first begins to fall from the tree. The second spraying should be done from one to two weeks after the last prunes have been harvested. If but one application is attempted, spray in the middle of September or when the maximum number of prunes are falling from the trees, as this would be the best time to kill the greatest number of larvae. The only satisfactory material to be used is a 12^ per cent solution of a high gravity miscible oil (33° Baimie). The proportions are as follows : Fart. Miscible oil (33° Baume) 1 Water 7 Put the requisite amount of oil in the pail or barrel to be used and add about one-fifth that amount of water. After some stirring this will become a thick creamy liquid, whereupon the remaining amount of water may be added with constant stirring. This should be continually agitated while being applied. The ordinary barrel or bucket pump will serve very well in apply- ing the spray to small trees. A good power apparatus, however, is necessary in order to compel the spray to reach to the top of large trees or to penetrate through the heavy foliage of thick hedges. Thoroughness of the application can not be overemphasized. It is absolutely necessary for successful control. See that the spray comes in contact with every twig and that all the larger limbs and trunks are thoroughly drenched. When planting trees not intended for trimmed hedges or wind- breaks, leave a wide space between each individual. It is a common fault to plant all sorts of trees too closely. Cypresses planted purely for ornament should be fully 40 or 50 feet apart. Trees already planted can be thinned out to this distance. This will retard the spread of the insect and give more nourishment to each tree left. The addition of fertilizers and water about the base of infested trees will also aid in overcoming the effects of the scale insect. In badly infested regions it is not advisable to replant cypresses. There are many other species of trees which are less prone to infesta- tion and are just as ornamental, which should be planted. There are other trees and plants, also, which make effective trimmed hedges. 22 BULLETIN 838, V. S. DEPARTMENT OF AGRICULTURE. The common privet (Ligtistnmi vulgare Linn.) forms an admirable hedge. The holly-leaf cherry {Primus ilicif&lia Walp.), Atriplex canescens James, and Pittosporwm spp. also are recommended. Pit- tosporum, however, is subject to attacks from scale insects which are just as difficult to control as the cypress bark scale. The Oriental and Italian cypresses form quite effective coniferous hedges, the lat- ter being tall and slender. Certain forms of red cedar {Jmiiperus virginiama Linn.) are used as trimmed hedges in certain sections of the United States. If this proves to be immune to the scale insect, it should be a very good substitute for the Monterey cypress. SUMMARY. The main cause of the browning and death of so many cypress trees, hedges, and windbreaks throughout California is the cypress bark scale, Ehrhomia cupressi. It was found in the course of a thorough investigation that the scale insect was not a native of the Monterey cypress, but of the incense cedar which occurs in the mountains of California, Nevada, and southern Oregon. From this host it has probably spread to the Monterey cypress by the transportation of incense-cedar seedlings or rustic timber to the regions infested. The characteristic injury caused by this insect begins to show on one or two limbs and slowly spreads to the rest of the tree. The foliage turns first yellow, then red or brown, giving the tree a very dilapidated appearance. After a few years the whole tree dies. The food plants of the cypress bark scale are Monterey cypress, Arizona cypress, Guadalupe cypress, and incense cedar. The larvae are small oval bodies, pale yellow in color, which are active for a short time after hatching. They attach themselves in crevices of the bark and are soon enveloped in a white cottony secre- tion. As they reach maturity they become reddish-brown in color and nearly spherical in shape. Oviposition begins in the spring and lasts throughout the summer. The eggs hatch into larvae in less than an hour and soon attach themselves. The females reach maturity in the fall and hibernate over the winter, starting oviposition in the spring. The males ap- pear in the late fall or early winter to mate and die. There are several insects which prey upon the cypress bark scale, none of which, however, is abundant enough to control the scale insect. Consequently remedial measures have to be adopted. A 12^ per cent solution of a high-gravity miscible oil is the spray recom- mended. To obtain complete control it is necessary to spray twice in the early fall, once in August and once in the latter part of September. UNITED STATES DEPARTMENT OF AGRICULTURE j>-^'^w<-> 1 BULLETIN No. 841 Contribution from the Bureau of Entomo'ogy L. O. HOWARD, Chief jTLiPf^'^'OT- Washington, D. C. PROFESSIONAL PAPER May 7, 1920 THE WESTERN GRASS-STEM SAWFLY By C. N. AiNSLiE,' Entomological Assistant, Cereal and Forage Insect Investigations CONTENTS Introduction 1 History 2 Food plants 8 The egg 9 Development ot the egg 10 The larva 11 The pupa 16 The adult 17 Oviposition 19 Key to North American species of Cephus. . . 22 Natural control 23 Artificial control 24 Cephus pygmaeus (L.) 26 Description -. 27 INTRODUCTION The western grass-stem sawfly (Cephus cinctus Norton) (fig. 1) is in many ways one of the most interesting and important insects that has attracted the especial attention of economic entomologists in recent years. It is a species native to the Laiited States and has been gradually coming into prominence since the beginning of the present centm-y by reason of the change which the feeding habits of the larvae have been undergoing subsequent to its discovery. Originally a grass feeder, it is becoming a serious menace to the grain growers of the Northwestern States because of its acquired appetite for small grains, within the stems of which it now subsists. Such changes of diet are probably occurring everywhere with greater frequency than formerly was deemed possible, especially among the phytophagous insects of the Middle West. When given a 1 The writer wishes to express his appreciation of the assistance afforded hy Messrs. J. C. Crawford, A. B. Gahan, and S. A. Rohwer, of the Bureau of Entomology, in the preparation of thi.s paper, the two former in determining parasitic material reared during the progress of the studies of the sawfly, the latter in making a critical examination of a large series of sawfly individuals reared or collected from various parts of North America, and for furnisliing detailed descriptions of Cephus cinctus besides a key to the North American species of the genus Cephus. Helpful criticisms from these men have added to the value and accuracy of the paper. The writer desires also to mention the valuable assistance and cordial cooperation of Mr. Norman Griddle, entomological field officer for Manitoba, Canada; of Mr. A. P. Henderson, of Bottineau, N. Dak., and of the several county agents in the infested areas, who have aided in various ways in the accumulation of infor- mation and material. 150056°— 20— Bull. 841 1 2 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. chance to ivod upon the various cultivated plants grown in bulk by the farmer or gardener, many of these insects gradually desert their native host plants and to a greater or less degree change their habits, including in their fare the more succulent and easily found food. HISTORY The existence of the western grass-stem sawfly v\^as first made known in 1890 when Mr. Albert Koebele reared adults from larvse that were mining in the stems of native grasses growing in the vicinity of Alameda, Calif. ^ During the next year, 1891, the species was described under the name of Cephus occidentalis by Messrs. Riley and Marlatt, from a series of individuals reared by Mr. Koebele and also Fig. 1.— Western grass-stem sawfly ( Cephas cinclux): Adult female. Much enlarged. from cotypes that had in the meantime been collected in Nevada and Montana.- In connection with this description the prophetic suggestion was made that: "The economic importance of this species arises from the fact that it may be expected at any time to abandon its natural food-plant in favor of the small grains, on which it can doubtless successfully develop." Nothing more was heard of this sawfly until 1895, when the late Dr. James Fletcher, Entomologist to the Dominion of Canada, swept adults at Indian Head, Northwest Territories, on July 5. He believed it to belong to the European species, Cephus pygmaeus 1j., and under 1 Koebele, A. Notes. In U. S. Dept. Agr. Div. Ent. Insect Life, v. 3, p. 71, 1890. 2 Riley, C. V., and Marlatt, C. L. Wheat and Grass Saw-Flies. In U. S. Dept. Agr. Div. Ent. Insect Life, V. 4, p. 168-179, 1891. (See p. 177-178.) THE WESTERN GRASS-STEM SAWFLY. 3 this name it was mentioned in his report for 1896^ with the further statement' that wheat straws containing Cephus larvae had been sent in hy Mr. John Wenman of Souris, Manitoba, who stated that the in- jury done by them was very slight. Nevertheless the prophecy of five years before had been fulfilled, since these grass feeders actually had attacked small grain. In 1902 Dr. Fletcher reported, in a private letter, that he had found the larvne nirmerous in grasses in the Northwest. In 1905 and 1906 Mr. G.I. Reeves, an agent of the Bureau of Entomology, noted the work of the larvae in various grasses, chiefly Agropyron sp., in Wyoming and the Dakotas, and in 1906 the same observer found the larvae attacking wheat sparingly near Kulm, N. Dak. Fig. 2. — Distribution of the western grass-stem sa\^'fly in the United States. August 31, 1907, Mr. E. O. G. KeUy, then an agent of this bureau, noted a few wheat straws near Minot, N. Dak., that had been bur- rowed by the larvae of Cephus. In 190S Messrs. F. M. Webster and G. I. Reeves found the larvae of Cephus working in grasses in the Willamette Valley in Oregon. In the same year Dr. Fletcher again called attention to this insect, stating that in the previous autumn it had appeared in central Manitoba and in the southeastern part of Saskatchewan in much more serious num- bers than ever before, and that the quantity of broken straws in the fields was causing the farmers some alarm. Mr. Norman Criddle of Aweme, Manitoba, a close observer and practical farmer, wrote to Dr. Fletcher that this fly had increased considerably during the last year or two, and was turning its attention to wheat and rye. > Fletcher, J. Report of the Entomologist and Botanist, 1896. Can. Dept. Agr. Exp. Farm, 1897. (See p. 229-230 ) 4 BULLETIN 841, U. S, DEPARTMENT OF AGRICULTURE. August 20, 1909, Prof. H. B. Penhallow reported from Sherwood, N. Dak., that he had examined about a hundred fields from Minot, N. Dak., north to the boundary line and several miles into Canada and had found larvae; present in every field but one. He estimated the damage in these fields as ranging from 5 to 25 per cent of the crop, but spoke of one field about 27 miles east of Sherwood where the damage was said to have exceeded 66 per cent. R. W. Sharpe reported similar damage in the Red River Valley, near Fargo, N. Dak. During 1911 and 1912 the writer found the species occurring freely in the native grasses in various parts of Utah, and as occasion offered Fig. 3.— Plants of Elymus condensatus growing along the railroad right of way. The natural habitat of the western grass-stem sawfly in Utah. the life histor}^ of Cephus was learned. Most of the facts in this paper are the result of this study. (Fig. 3.) During the years 1913, 1914, and 1915 the writer has found this sawfly almost universally distributed over the Dakotas, IVIinnesota, Iowa, and Nebraska, feeding in Elymus, timothy, and Agropyron at Elk Point, S. Dak., in Agropyron tenerum, near Chamberlain, S. Dak., in timothy at Edgeley, N, Dak., in Bromus in^rmis near Merricourt, N. Dak., in Elymus canadensis at Shakopee, Minn., in practically all these grasses near Sioux City, Iowa, and in wheat, timothy, and Elymus near Minot, N. Dak. It seems to have little choice in the various native grasses and is ready to attack any of the cultivated THE WESTERN GRASS-STEM SAWFLY. 5 sorts provided the stem is sufRcientlj large for the larval gallery. As a rule, the larger, more robust stems are chosen for attack, espe- cially in cultivated grasses such as timothy and Bromus. Blue gi-ass and similar slender-stemmed species appear to be immune. It is a httle surprising that a minute examination of Stipa viridula from New Mexico developed the fact that none of the stems of this robust grass were infested. This Stipa was gathered in northern New Mexico, gi'owing in almost the same latitude as the Elymus condensatus near Pinto, Utah, where the fly abounds. August 25, 1916, the writer, then at Pierre, S. Dak., received instructions from the Bureau of Entomology to visit Bottineau County r , -;.^«^^2^. ^^^^^^^H^l Fig. 4,— Wheat lield of Thomas Yearn, near 6oiiris, N. Dak., shuwiug heti\ > uinnuge done by the western grass-stem sawfly in 1916. in North Dakota and investigate injury to wheat. It was believed locally that the Hessian fly was responsible for the damage that was being done. A very superficial examination of the injured fields proved beyond a doubt that the Cephus was present in large numbers and was doing an inuuense amount of mischief. Every field was infested, not only in Bottineau County, but in the adjoinmg counties of Benson, Pierce, McHenry, and Rolette. Near Souris, a few miles south of the Canada line, a large field of wheat on the farm of Thomas Yearn was fairly carpeted with the ' ' straw-f alien " grain. (See figs. 4 and 5.) The loss from Cephus injury in this field was estimated at 60 per cent or more. Six feet of drill row here were taken at random and examined plant by plant. Forty-eight infested stubs were found. 6 BULLETIN 841, V. S. DEPARTMENT OP AGRICULTURE, an average of eight to each foot of drill row, This would mean 150 to the square yard or about 726,000 larvae to the acre. Higher counts were made later in this same field, so the average may be larger than sta- ted. During April, 1917, IMr. Yeam's field was again visited and a ran- dom square yard marked out and counted. Two hundred and sixty-nine infested stubs were taken from this yard, which would mean more than 1,300,000 larvse to the acre. Fifty of these stubs were opened and 47 of the imprisoned larvae that had spent the winter within the straw were found to be normal and very much alive. The proportion of living individuals among the hibernating larvae seldom falls below this ratio. Fig. 5.— Bird's-eye view of wheat in Thomas Yeam's field, Soiiris, N. Dak. Ninety per cent of these fallen wheat stems have been mined by the western grass-stem sawfly. The spring of 1917 witnessed a peculiar condition of tilings in Bottineau and the adjoining counties of North Dakota. The dry weather hindered the growth of both grasses and grains, so that when the adult Cephus began to appear in June there was almost no oppor- portunity for oviposition. Stems of Bromus from chance sods grow- ing among wheat and on waste ground were filled with eggs. Young plants of spring wheat that had barely begun to joint were attacked and often contained as many as tliree and four eggs placed in the stem close to the ground. With a few strokes of the net 136 adults were swept from young wheat, so numerous were the flies at that time. In spite of the unfavorable oviposition conditions of that spring, the THE WESTERN GRASS-STEM SAWFLY. 7 eggs appear to have hatched and at harvest time the majority of the wheat stems had been bored and many were cut off at the base. Careful harvesting and the use of horse rakes saved a large proportion of what otherwise would have been a total loss. The infestation was much more general than in 1916. A somewhat hasty reconnaissance was made through north-central North Dakota in August, 1919, that it might be ascertained as definitely as possible just how the Cephus attack was progressing. A number of fields in Bottineau County were examined and found to be heavily infested. ^lost of these had been raked after harvest and it was consequently impossible to compute accurately the percentage of infestation. The numerous sawfiy-inhabited stubs in the drill rows, however, proved the severity of the attack. It was roughly estimated that about 30 per cent of the grain had gone down in most of these fields as the result of Cephus work. This figure is probably very conservative. It is conceded by many observers in that region that the injury during the year 1919 was greater than during any ]>revious year since the study of this pest was begun. More fields had been seriously invaded and were injured to a larger extent than had before been observed. Even fields of durum wheat, hitherto believed to be nearly free from fly attack, were severely injured in 1919, if the statements of reliable farmers are to be accepted. The question of immunity of durum wheat will be discussed later in this paper. It may be stated, however, that the farmers are profiting by past experience and have used horserakes in stubble fields to such an extent that the percentage of actual loss of grain has been reduced to a small figure. The quality of grain from the fallen straw is naturally somewhat below the normal, since the work of the larvae in the stems produces some injury in the heads as they fill. Cephus was found mining wheat near Hettinger in southwestern North Dakota, July 18, 1917. September 22, 1917, infested wheat was found near Mott, 30 miles north of Hettinger. In October of the same year many wheat fields in Towner and Cavalier Counties, in northeastern North Dakota, showed heavy infestation, although dur- ing the previous year it was difficult to discover more than a trace of Cephus presence in the wheat in tliis region. None was found in the vicinity of Fargo, although it doubtless occurs throughout the entire Red River valley. A gathering of sods of Elymus canadensis sent to the writer from Charleston, Mo., during the summer of 1917 contained at least one larva of Cephus cinctus that had been boring the stem of this grass in that region. This locality is a little south of the latitude of Pinto, Utah, where this insect abounds. 8 BULLETIN 841, V. S. DEPARTMENT OF AGRICULTURE. Roughly speaking, so far as is now known, the sawfly inhabits an area bounded on the north by a line far into Canada; on the east by the Mississippi River, or probably a little east of that; on the south by latitude 36°; and by the Pacific Ocean on the west. From the foregoing brief summary of its history it will be seen that CepJius cinctus is distributed over an immense territory and that it constitutes a potential menace to the small grains throughout tliis vast area. As the acreage of native grasses is decreased from year to year by the bringing of wild lands under the plow, pests such as the sawfly will be forced to depend in an increasingly large measure upon the small grains and other products of the farms. On this account the injury caused by these formerly harmless insects bids fair to increase steadily. In the past the numbers of grass-feeding insects such as the one considered in this paper have been governed mainly by the supply of food plants. A dry summer that retarded the growth of long-stemmed grasses would automatically reduce the numbers of the insects that lived within these grass stems and perhaps bring certain species to the point of extinction. It is easy to see how seasonal fluctuations in vegetation would, to a large extent, either multiply or diminish the numbers of these insects. Then again, the farmer, by introducing fields of grain into a region previously uncultivated, brings in conditions unknown before and invites the attack of these and other formerly harmless insects, making it possible for them to become a menace to his future. Such a study of life history as has been attempted in this paper is urgently necessary in order that control measures may be undertaken suc- cessfully when such insects become pests. FOOD PLANTS The various species of Agropyron and Elymus, genera both of which are well represented in the West, appear to have been the original hosts of the larv?e. Since their feeding habits have been modified by changing agricultural conditions, the list of their present host plants, so far as known, stands as f oUows : Elymus canadensis Agropyron occidentale Calamagrostis spp Elymus condensatus Agropyron caninum Festuca sp. Agropyron tencrum Hordeum jubatum Wheat Agropyron richardsoni Bromus inermis Durum Agropyron smithii Phleum pratense Spelt Agropyron repens Deschapsia sp. Rye Barley probably should be added to this list. Since the larva is wholly unable to move from one stem to another, it is very obvious that the host stem must be large enough to afford both shelter and food during its entire growing period. Hence only the larger-stemmed grasses can be mined successfully by the Cephus THE WESTERN GRASS-STEM SAWFLY. 9 larvse. Occasionally an unusually vigorous plant of a slender- stemmed grass, like Hordeum juhatum, affords stalks with diameter sufficiently great to be attacked by Cephus. Small grains, such as wheat and rye, readily serve as hosts to this insect, because they are of suitable size and the length of their grow- ing season coincides with the growth of the larva. Even if harvest time should happen to come before the maturity of the larva, the reaping machine probably would sever the stem far enough above ground to leave the larva below the sickte cut, where it could house itself safely before the end of the season. Judging the future by the recent past, it seems probable that this fly, before another decade is past, will be found attacking practically all of our native and cultivated grasses and most of our grains. It must be remarked in this connection that up to the present time this species has confined itself entirely to the West and has been found in only a few localities east of the Mississippi River. Its choice of wheat for food has taken place, so far as known, only in North Dakota and western Canada, although it is probable that Montana wheat fields have been invaded. From present appearances its attacks probably will be confined to vegetation growing within the area where spring wheat is sown. THE EGG The egg of Ceylius cinctus is, when newly laid, decidedly crescent- shaped, glassy in appearance, milky-white in color, usually quite symmetrical, the ends of the crescent tapering and rounded. It is marked by very faint, short, longitudinal lines or wrinkles, placed without regard to order or pattern. The size of the egg varies with the size of the female that pro- duced it and measures from 1 mm. to 1.25 mm. in length. The greatest breadth is about one-third the length. The covering membrane is hyaline and transparent. Although very tliin and dehcate it is sufficiently strong so that the egg may be safely lifted and moved by the aid of a fine brush. The egg always lies free witliin the stem of the host plant, either in the stem cavity or in a hollow excavated by the ovipositor of the female that placed it. This cell is always a little larger than the egg, so that it is com- paratively an easy matter to remove the egg to a moist cell or else- where for study. The number of eggs distributed by each female appears to vary but little. Dissections of a number of adults taken in the field and of others reared in captivity agree in most cases in giving a count of about 50 eggs in the ovaries, these eggs being, as a rule, equal in size and apparent maturity. 150056°— 20— Bull. 841 2 10 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. DEVELOPMENT OF THE EGG After a number of trials it was found to be impracticable to rear the egg in situ, since it was next to impossible to maintain the proper moisture conditions within the stem. The method that finally was adopted, and that gave excellent results, was to remove the egg from the stem and place it in a minute drop of water within a small thin watch glass which was then immediately inverted on a glass slip and sealed with a ring of water to prevent undue evaporation. This form of moist cell proved quite satisfactory and permitted con- tinuous examination of the egg with a moderately high-power lens during the entire period of incubation. It was found necessary, in order to continue the requisite moisture supply during a period of several days, to invert over the sealed cell a larger watch glass and over this in turn a tumbler. In this manner evaporation was re- duced to a minimum. It is altogether probable that the amount of moisture in such a protected cell exceeded that normally present within the grass stem, but in every egg treated in this way the incu- bation appeared to proceed naturally. Temperature and moisture are, without any doubt, the prime factors that hasten or retard the egg development. The temperature maintained within the laboratory dm^ing the course of these inves- tigations was much more equable than that in the field, where, as in Utah, the heat of the sun through the daytime, followed by a cliilly night, must alternately hasten and check development. The data given below, therefore, may only approximate what actually takes place under field conditions. A few hours after the egg leaves the oviduct the milky-white con- tents of the egg which at first completely filled the envelope slu-ink a little from each end leaving a transparent space or vacuole. Grad- ually the interior mass of exceedingly minute particles coalesces until about the second day when a series of faintly discernible cells arranging themselves along a central axis begins to appear. Early on the third day the form of the larva can be dimly seen, the head being almost transparent and filling one end of the egg sac. The body is looped on itself, the cauda folded beneath the abdomen and extending forward nearly to the head. By the close of the third day the abdominal segments are usually well defined. During the fom^th day, in most cases, a spasmodic and intermit- tent heart beat may be noticed. These pulsations become more and more regular as the houre pass and during the fifth and sixth days the heart beats with much regularity at the rate of about 120 im- pulses per minute. At intervals, for .some unknown reason, it may slow down to 75 beats, but soon resumes its former rate. The head appears abnormally large at this time, but although its general outlines are well defined the brown jaws and eye spots are THE WESTERN GRASS-STEM SAWFLY. 11 not yet visible. Over night, at the close of the fifth day, the jaws turn brown and the eye spots appear and darken. Usually, after the fourth day, the muscular system of the larva is in almost con- stant motion, sliifting and adjusting, with the heart pulsating and the muscles moving, all clearly to be seen through the transparent membrane that serves as the shell. The activity of the larva within the sac increases during the sixth day, and either on this day or the seventh it escapes from its confine- ment by a series of convulsive movements that rupture the delicate shell and set it free. After the first day the egg changes shape, becomes intumescent, generally loses its crescentic shape entirely, and grows oval or reni- form in outline. THE LARVA When it escapes from the egg the larva ffig. 6) possesses a very large head armed with a pair of powerful biting jaws, a weak, slender body, and a most vigorous appetite. It is very active from the start and begins almost at once to feed upon the living parenchymatous tissue by which it is surrounded in the inte- rior of the stem, excavating for itself a threadlike gallery both above arid below the spot where the egg formerly lay. The larva is at first nearly transparent and colorless until it becomes filled with the tissue on which it exists. The body segments are strongly and clearly marked from the time the larva leaves the egg. The jaws are brown, three or four Fig. 6.— western grass-stem sawfly: . , 1,1 • , I'll 11 11 Newly-hatched larva. Greatly pointed, the points chisel-shaped, beveleci enlarged. on the inside edge. The brown face plate is filled with crossed bands of striated muscular fiber that actuate the powerful jaws which form the most important item of the domestic economy of the young Cephus. The caudal horn, by means of wliich the larva moves up and down in its gallery, is also brown and is armed, even in the first instar, with a series of stout bristles at the base of its cylindrical and squarely truncate extremity. The larva is footless, the position of the legs being marked by minute, rounded tubercles terminating in a few short bristles. Although the primary excavation made by the larva may extend for a short distance above the egg cell, the general course of the progress is invariably downward. In its earlier stages of existence, at least, the larva traverses its gallery several times, swallowing repeatedly the same fragments of tissue that have already been devoured during the first excavation of the stem. Young larvsB are 12 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. frequently found several inches above the lower end of the boring, moving through the solidly packed "sawdust." As the larva ap- proaches maturity it is doubtful if it ventures into the upper and slender part of the stem, but it still reworks the frass farther down, enlarging the bore in places. The number of instars is difficult to determine, owing to the larval practice, just referred to, of passing all the frass several times through the digestive apparatus. Nearly all the cast skins disappear com- pletely under tliis treatment, only the heavily chitinizod parts such as the jaws and caudal horn being recognizable in the burrow. Care- FiG. 7. — Western grass-stem sawfly: Mature larvae removed from their galleries. Enlarged 4 diameters. ful investigation of these fragmentary remains appears to establish the fact that there are four molts. The contents of innmnerable stems have been examined with scrupulous care and with varying results. In a few cases as many as four sets of jaws and in others four caudal horns have been found, mixed with the frass within the stems. Seldom were more than four sets removed from a single stem; usually only three were found. As is stated elsewhere in this paper, it is no uncommon tiling to discover two and even three larvae mining a single stem, although but a single individual can possibly reach maturity with the amount of nutriment contained in THE WESTERN GRASS-STEM SAWFLY. 13 one stem. It is believed that the larva that finally reaches maturity has devoured its rivals. It is obvious that the remains of these superfluous individuals would naturally be counted when a census of exuviae was undertaken and would complicate the result. But from the best evidence obtainable it is almost certain that there are five instars in the larval life of this species. The length of the larval period is probably about 60 days, varying more or less with the warmth of the summer and the state of maturity of the host stems. The acceleration or retardation of the oviposi- tion period owing to an early or late spring has much to do with the date of maturity of the larvae, and possibly with the length of the larval period. August 29, 1911, at Kimballs, Utah, at an elevation of 7,000 feet, the writer found mature larvae in stems of Ehjmus condensatus. The next year, at the same place, oviposition was beginning freely during the first week of 'Tuly. The determination of the larval period is wholly inferential, based upon the find- ings in a series of stems (figs. 7 and 8). The full-grown larva? vary greatly in size, their growth being governed, as is usual in the case of such borers, by the ({uality and quantity of food consumed. Those living in wheat stems are much smaller as a rule than those found in rank- growing grasses such as Elymus. Measurements of a series of individuals give variations of from S to 14 mm. in length and from 1 to 2 mm. in diameter. When mature the larva always seeks the ex- treme base of the stem, where it soon begins its preparations for hibernation. Its first move is to cut a neat V-shaped groove entirely around and inside the stem, usually at or a little above ground level. This groove never severs the stem completely, but so weakens it that the upper stalk, swayed by the wind, will break off com- pletely when dry, leaving a stub that is very characteristic of the work of this insect (fig. 5). In tliis simple manner the larva provides for the easy escape of the adult in the following summer. The length of the stub thus formed varies greatly. In Elymus condensatus the stub sometimes will project above the ground as much as 3 or 4 inches, wliile in other grasses, and especially in wheat, stubs easily can be found less than an inch in length in all. Instances have been observed where two or more grooves had been cut inside the same stem, as if the larva had been uncertain as to the best place for severing the grass. After cutting its characteristic groove within the stem the larva forces a mass of the debris into the Fig. 8. — Western grass-stem sa^\'fly : Mature larva. Enlarged 5 diameters. 14 BULLETIN '841, V. S. DEPARTMENT OF ACxRIC'ULTURE. bore just below the groove and in this manner plugs the upper end of the stub that is to be left in the ground after the upper stalk has been broken away (figs. 9, 10, and 11), This dry frass in some manner is packed firmly into its place, perhaps by means of pressure rather than by being cemented with a liquid furnished by the larva, since the plug is readily penetrated by moisture. This is somewhat remarkable in view of the fact that an undue amount of moisture appears to have a disastrous effect upon the mature larva. One would suppose that these stubs, often wholly submerged in water- soaked earth for weeks at a time, would absorb, during the long period of liittemation, a fatal amount of dampness from the rain or melting snow. But there is no evidence that this ever happens. Fig. 9. — Western grass-stem sawfly: Infested wheat stubs from Bottineau County, N. Dak. September 16, 1911, one of the larvae was removed from the hiber- nation chamber and placed in a small vial, still inclosed within the silken tube or cocoon, which was unbroken. For months this larva remained passive and motionless except when the vial was exposed to bright sunshine. Because of the light or heat, or both, when placed in the sunlight it would become active at once, and travel up and down within its cocoon in its efforts to escape. January 20, 1912, to prevent the air in the vial from becoming too dry a small drop of water was introduced and the vial again corked tightly. An hoiu" later it was noticed that the silk tube had collapsed and the larva within was limp and apparently dying. The surplus moisture was removed quickly, whereupon the larva revived almost at once. THE WESTERN GRASS-STEM SAWELY. 15 If the same amount of moisture had entered the stem where the larva was hibernating it probably would have caused its death. This experiment, taken in connection with others that were not so directly conclusive, seems to prove that the porous plug in the stub must prevent in some way the admission of an undue amount of moisture into the chamber below, although water readily penetrates it. The gallery below the plug is always entirely free from debris, forming a hibernation chamber and later a pupation cell. Within this chamber the larva lies with its head up and usually pressed against the barrier at the top, always on the alert to retreat down- ward at any sign of disturbance. It descends l)y alternately flexing and straightening the body, bracing itself first by the jaws, then by the caudal horn as it hitches its way down In ascending, the caud'?J horn is thrust against the side of the gallery or the cocoon, the body is straightened, the jaws obtain a pur- chase to hold the dis- tance gained, when the body is again drawn up until the caudal horn is applied to the side wall for another push. Late in the summer or during the autumn the larva spins for it- self within the hiber- nation chamber an almost transparent tube of filmy silken tissue. This silk tube is sometimes several times the length of the larva, is closed at both ends, and is free from the sides of the chamber, so that often it can be readily withdrawn entire. When first constructed this fabric is comparatively strong and pliant but after some months it grows more brittle and is easily ruptured. As a rule it remains intact until the emergence of the adult. Even the presence of half a score of parasitic larvae often fails to wreck the deli- cate structure during the winter. The longevity of the sawfiy larvae is remarkable and is worthy of mention. September 8, 1911, a number of stubs of Elymus conden- satus containing Cephus larvae were gathered and set upright in sand within doors. From time to time this sand was moistened but finally was allowed to stand perfectly dry. During October, 1912, Fig. 10.— Western grass-stem sawfly: Infested wheat stubs, enlarged 3 diameters, the two left-hand ones opened to show hi'iernating lar\';T> in silii. 16 BULLETIN 841, U, S. DEPARTMENT OF AGRTCIULTURE. these stubs were examined and a number of the inclosed larvae were found to be still living, active, and unchanged. Four months later, 17 months from the time they were gathered, they were still alive and feebly active. Infested stubs of the same grass taken during September, 1912, and treated in the same manner, contained at least one living larva on February 23, 1916, 3 years and 5 months later. The others had nearly all died within about 30 months of the time they were gathered. It is possible that the lack of necessary moisture may account for the retardation of these captives. However, the same retardation of development has been noted in the field. Inhab- ited stubs of the previous year's growth of grass and grain not infre- quently have been found, containing larvae that were to all appearances entirely normal and active. It ap- pears more than probable that in this manner the perpetuation of the spe- cies is assured in case of unfavorable seasons. During the winter the larvae are, of course, frozen, or are chilled into immobility and show no signs of life when disturbed. As soon as the earth warms in the spring they again grow active and move freely up and down within the limits of the silk-lined hibernation chamber until the time of pupation arrives. THE PUPA The pupa when first formed is riG. 11.— stems of wheat grooved internally, milk-white, slender, and somewhat longer than the larva from which it was derived. Its average length is not far from 12 mm. and its breadth is about 1.5 mm. At first the pupa lies motionless within the silken pupation chamber or cocoon (fig. 12) for probably a day or two, after which inactivity it again becomes animated. When disturbed it will endeavor to escape the threatened danger by moving either up or down the tube, hitch- ing itself along in much the same manner as the larva but going a lesser distance with each effort. Like the Larva it is almost always found with its head pressed closely against the plug of frass at the upper end of the chamber. In a few cases pupae have been discovered heading: downward in the stem. It is doubtful if these can reverse by larvfe of the western grass-stem saw- fly. THE WESTERN GRASS-STEM SAWFLY. 17 their position, but the adults which issue are probably agile enough to turn about and escape. The duration of the pupal period is not known certainly, but is believed to be very brief, not more than a week at the most. After the first day the legs and body darken until they ])ecome a lustrous black within the transparent, almost invisible filmy membrane in which they are inclosed. This membrane is often lacking and may occasionally be destroyed by the movements of the pupa within the chamber. When fully mature the pupa changes within the cell to an active adult. This adult remains imprisoned until some unknown impulse compels it to force its way upward through the plug of frass placed at the upper end of the chamber by the larva 9 months before. The writer, by splitting stubs of grass or grain in June, has liberated adults repeatedly, which, when free, were able to take instantly to wing without any preliminary process of drying or other preparation. These adults were evi- dently resting, in perfect condition, waiting for some secret signal from the outside world before taking the final step for liberation. A very few die within the cell, possibly because of lack of vitality needed to break through the stopper of frass above them. In cases where the girdling of the stem was inefficiently done, so that the grass stalk did not break off during the winter season, the adult dies as a matter of course, since these flies apparently are not fitted with iaws capable of biting through the woody no- i2.-western grass-stem „ , ° ^ '' sawfly: Pupa. Three and stems 01 dry grass. one-halt times natural size. THE ADULT The adult Cephus ductus is a beautiful insect with a polished black body marked by three prominent yellow bands across the abdomen. The legs are yellow and the wings smoke-colored. The description, by S. A. Rohwer, follows: Length 7 to 12 mm. Head shining, polished ; anterior margin of clypeus truncate with angles prominent and sometimes slightly denticulate; antennae usual for the genus; thorax shining but with setigerous punctures on scutum; sheath nearly paral- lel-sided but a little broader at base, apex truncate with corners rounded; hypopy- gidium rather narrowly subtruncate apically. Black marked with bright lemon yellow, amount and extent of yellow markings varying greatly; head of female usually black but more rarely with face entirely yellow or having yellow spots; head of male black but always with yellow on face; thorax black, the upper angle of mesepister- num, parapteron, and scutellum (usually) yellow; legs yellow with coxse, trochanters (occasionally both of these having yellow marks), bases of femora more or less, apices of tijiie and tarsi sometimes, black; hind tibiae and tarsi sometimes reddish yellow; abdomen black, spot or band on second tergite, band on third, fifth, sixth, and eighth 18 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. tergites and lateral margins of tergites yellow, the size and extent of these markings varying and occasionally the fourth tergite having a yellow band ; wings fuliginous, venation dark brown, costa and stigma yellow. The female is noticeably larger than the male, and in the field is captured much more easily. The characteristic attitude of the adults of either sex while at rest is to lie fiat against the grass stem, head downward, the body closely appressed to the stem, the legs not spread but stretched in line with the l)ody while the body itself is concealed behind the closely folded, smoke-colored wings. The ease with which such a strikingly colored fly, while in this position, can escape observation, is remarkable. During the chill of the morning and after sundown this attitude is universally assumed. When basking, in the sun at midday, on the warm side of a grass stem, the fly is much less compact, with the wings partly spread and the legs outstretched in order to absorl) the utmost of the warmth. Like most Hymenoptera, this species is very partial to sunshine and rarely is seen abroad on a cloudy day. In fact, in cloudy weather it is not easy to find these flies at all, unless one is entirely familiar with their habits. They are weak fliers and seldom travel to any great distance at one time. In Utah they commonly move about among the plants of bunch grass, making short flights from tuft to tuft. If the wind rises or the sun goes behind a cloud they promptly disappear until conditions again become satisfactory. The writer has never taken the adults at any great distance from their breeding places. Their hovering flight is peculiar, the swaying motion of their bodies in the air reminding one of certain tipulid flies during their mating air dance. They often hover for a long time to the windward of a grass plant without alighting, seeming to enjoy the motion. The males are on the wing much more than the females, but neither sex will remain in the air while the wind is strong or when it is cool. The adults are not at all timid and can often be readily taken from the grass stems with the fingers. When conditions are favorable for her the female is usually too intent on oviposition to be easily annoyed but if disturbed beyond endurance she quickly disappears, her dark color and slender body enabling her to vanish completely among the vegetation. Copulation is very brief, usually lasting less than a minute. No notes were made on the attitude assumed durmg the operation. The species is single brooded, the adults appearmg during the spring and going out of existence some time about midsummer. The earliest individual met in Utah was taken in a net April 26, 1910, in an alfalfa field. Adults have been seen in the mountains late in July and they probably linger longer than that, ovipositing in such green grass stems as they can find. Near Kimballs, in Utah, September 8, 1911, the writer took very young larvje from stems of THE WESTERN GRASS-STEM SAWFLY, 19 Elymus condensatus, growing from plants that had been browsed by cattle and had thrown up fresh green stalks. Mr. Norman Criddle states that in Canada the adults appear during the second week in June and may be met with until about July 10. Occasionally they may be found feeding on flowers. Doctor Fletcher has taken them in Canada on flowers of the tumbling mustard. It is unquestionably true that the time of their appearance and the length of adult life are both largely governed by climatic influences and vary with the season. When confined in emergence tubes or other limited places the males develop savage instincts and attack each other without mercy, usino- their jaws freely to snip off the antenna?, and, in some cases, the legs of their rivals. Singularly, very few of the females confined with them are thus mutilated. OVIPOSITION Weather conditions have always been an important factor in con- trolling the oviposition of very many of the Hymenoptera, and they are of particular importance in the case of the Cephus. These flies go into hiding when the day is dark, damp, cool, or windy. Only on bright, warm, still days are they to be found busy with the operation of placing their eggs. In Utah, where the first studies of their habits were made, the mornings and evenings are chilly as a rule, hence the activity of these flies is confined to the hours near midday. They are every^vhere the most active between the hours of 10 a. m. and 2 p. m. The swaying of the grass and grain stems in the wind appears to be a hindrance to them in alighting and ovipositing. A gentle breeze will often keep them hovering for several minutes to the windward of their goal, while a sudden mountain gust is apt to put an abrupt end to all efforts for the balance of the day. Their actions are con- trolled by unknown factors, for sometimes on a still, sunny day they will spend much of the time roosting on the stems, while again, under apparently the same conditions, they are constantly in motion, flying and hovering a long time before alighting. While the female is poised in the air before a sod of grass or grain she is evidently busy selecting the particular stem in which she hopes to oviposit. Once she has chosen and settled, she seldom changes to another stalk, although she may halt at several places on a single stem and attempt oviposition at each pause. Occasionally, after a hasty examination, she may again take to wing and make another choice. Repeated observations seem to have established the fact that one of the chief requisites of a proper stem is that it shall not yet have put forth a head. In all the countless instances where ovipo- sition has been observed, the female has never been known to choose a stem with a head. 20 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. When she has made her selection of a suitable stem, the female usually alights about half way up and runs briskly to the upper end, halting almost imperceptibly every few steps. The gait of an ascend- ing fly is so characteristic that it determines with much certainty if the individual is a female intent on oviposition. Arriving at the apex of the stem, after a care- ful survey of its condition, she frequently makes an elaborate toilet, preening herself most care- fully, until she is in perfect condition. She then descends, exaggerating slightly the hesitating step by which she had ascended. The antennae are held horizontally in front of the head as she moves, and she occasionally touches the surface of the stem with their tips. There is none of the rapid anten- nal vibration so common among the smaller chalcids and many other Hymenop- tera. She gives no evidence of being in search of any particular point, but goes straight down the stem. When satisfied that she has gone far enough she halts abruptly, usually an inch or less above the second node from the top of the stem, slowly arches her abdomen and clasping her hind pair of feet around the stem as far as they will reach begins to drive the saws into the hard outer tissue. Figure 13 shows the attitude taken at this time. These saws are exquisitely fashioned, curved like a scimitar, double, very thin with serrated edges. (See fig. 14.) They are used to split the outer coating of the stem rather than to cut it, and they make an opening so exceedingly small that it is almost impossible to find the scar after the wound has healed. These saws are gradually" forced into the stem, the operation occupying a minute or more. In the field the female always heads downward during oviposition and the curve of the saw blades brings them, when fully inserted, in a line parallel to the axis of the stem. They are frequently partly withdrawn and then direction slightly changed. When the stem is in proper condition the saws are thrust in several times, as far as they will go, then are withdrawn, Fig. 13. — Western grass-stem sawfly: Female oviposit- ing. About life size. Fig. 14.— Western grass-stem sa^vtly: Saws, greatly en- larged. THE WESTERN GRASS-STEM SAWFLY. -21 the dorsal part of the pygidium being used as a fulcrum to extract them. They are inserted again, this time often with a twisting mo- tion as if trying to enlarge the opening. They are finally forced in as far as possible, as is evidenced by the tenseness of the rear legs straining at the stem, and are held in this position for half a minute or more. This is probably when the egg is deposited, the insect standing practically motionless except for a slight vibration of the antennae. A laboratory note may be of mterest, giving in detail some of the facts that have been mentioned above : June 5, 1912. The adult Cephus were emerging freely from, the Elymus material brought from the mountains, and it occurred to me it might be possible to secure some views with the camera if they could be induced to oviposit. Several females were ob- served attempting oviposition in the dry stems from which they had recently emerged. A green stem of Elymus was planted in a tumbler of wet sand and the camera focussed on this stem about midway. After a few trials I discovered that this stem must be short and headless. The females were taken from the cage and placed, one at a time, on the damp sand in the tumbler. Their first act in every case was to spend a long while drinking eagerly of the water held in suspension by the sand. A few of them sipped water for as much as half an hour before they could be induced to leave. WTien guided to the base of the Elymus stem they would usually ascend without a moment's hesitation. Once started they would go to the very top and there would either preen themselves interminably, or would wheel and descend with the usual cautious, hesitating gait, a few steps at a time. When part way down, without apparently choosing any especially suitable spot, the abdomen would arch and oviposition would begin. Sometimes these efforts were plainly failures, but some of the flies would sink their saws well into the tissue of the stem and stand for a number of seconds motionless, thus affording an opportunity for the camera. Much time was lost to-day because of the exasperating neatness of these insects. Each one would brush herself over and over again with the most minute exactness and no amount of urging would avail to shorten the process. The same careful preening has been frequently observed also in the field. Several life-size views of these flies were obtained to-day by the above method, views that would be impossible in the field because of the almost constant motion of the limber grass stems. Several of the females became confused to-day when compelled to remain on a certain part of the grass stem during oviposition, and faced up the stem instead of down as they invariably do, normally. When busy with oviposition they seem oblivious to whatever is gomg on around them, and the writer has repeatedly watched, through a half-mch triplet, the female manipulating her saws. Close observa- tion did not annoy her in the least when the lens was carefully handled, and she paid no attention to the proximity of the onlooker. Under the closest scrutiny it is impossible to determine just when the egg is passed into the stem. It is probably at the time when the female stands motionless after the saws have been driven in to their full length. The function of these saws appears to be twofold. At Pinto, Utah, in June, 1912, the writer found that the eggs were invariably placed 22 BULLETIN 841, t\ S. DEPARTMENT OF AGEItJULTURE. in a cell hollowed in the solid parenchyma of the stem of Elymus condensatus , this cell being a little larger than the egg. Besides piercing the stem, the saws are also of use in excavating this egg cell, in case such a cell is needed. At Kimballs near Salt Lake City, in the same grass, the eggs were nearly always placed in the hollow part of the stem, lying free in the central cavity. Normally but one egg is placed in each stem. However, no atten- tion is paid to previous oviposition and as many as five eggs have been taken from a single stem. As is stated elsewhere, only one of these larvsB can possibly survive until fall, so this multiplication of eggs simply means economic waste for the Cephus. The date of ovipo- sition varies with the latitude and the alti- tude. At Pinto, Utah, on the edge of the des- ert country and with a low altitude, newly hatched larvae were found June 14, 1912, while at Kimballs, 350 miles north of Pinto and with an altitude of 7,000 feet, oviposi- tion was beginning durmg the first week of July in the same year. Mr. Criddle states that in Canada most of the eggs are depos- ited during June. The date of oviposition in the Dakotas and in Minnesota is unknown. (Fig. 15.) KEY TO THE NORTH AMERICAN SPECIES OF CEPHUS Through the courtesy of Mr. S. A. Rohwer of the Bureau of Ento- mology a key for the determination of the known species of the genus Cephus occurring in North America is here presented. Stigma and coyta dark brown of a uniform color ; mesepisternum black ; femora black ; apical tergite and venter black; face and scutellum black (face of male with yellow spots) pygmaeus Linnaeus. Stigma in greater part and costa yellow; mesepisternum with the upper angle yel- low; apical tergite and usually the venter in part yellow; femora usually mostly yellow ; face and scutellum of female usually black but occasionally with yellow spots cinctus Norton. Fig. 15. — Life-history diagram of the western grass-stem sawfly. THE WESTERN GRASS-STEM SAWFLY. 23 NATURAL CONTROL In the usual scheme of things an undue increase of insect pests is controlled naturally by parasites that take a heavy toll of their hosts and prevent their multiplication. Under normal conditions, when the Cephus cinctus existed wholly in grass stems, the larvae were at- tacked with varying success by two or more species of parasites that destroyed numbers of them and kept them within reasonable bounds. Since the fly has begun to change its habits and to subsist on wheat and other small grains to a certain extent, these parasites apparently have not yet learned of the change and are confining their attacks, as heretofore, almost en- tirely to those larvae that they find in grass stems. A very few parasites have been taken from infested wheat stubble, and there is little question but that in course of time the busy little parasites will hunt their prey in the grain stems and do their share in helping to control this pest. The most common parasite found every- where in the grasses is Pleurotroins utah- ensis Cwfd., a beautiful little bronze-green chalcid that was reared by the writer from numerous larvae taken near Salt Lake City, Utah, from Cephus hibernation cells. This species appears to kill the larva only after it has formed its hibernation cell. It is gregarious and seldom or never attacks its host singly. As many as 12 of its larvae have been taken from a single cell, but 5 or 6 is a more common number. These larvae are white and measm'e from 2.5 mm. to 3.5 mm. in length. They are somewhat active and travel slowly about the cell when matm-e. They are often found crowded together in one end of the cell, but when distm-bed will scatter about the chamber (fig. 16). Although this species is widely distributed and propagates in numbers it appears to destroy but a small percentage, possibly 10 per cent of the Cephus larvae in the native grasses of Utah. In Bottineau County, N. Dak., it attacks the sawfly very freely in Bromus and timothy, and in some localities has killed more than 50 per cent of the Cephus larvae. Indeed, it and one other parasitic Fig. 16.— Larvae of Pleurotropis utah- ensis, a parasite of the western grass-stem sawfly, in sit7t. 24 BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. species are so numerous in these roadside grasses that it would seem poor poUcy to recommend the cutting of the grasses in midsummer as a measure of Cephus control. As has heen stated, few parasites have been found in stems of wheat, but, without doubt, they will learn very soon of the presence of Cephus in grain fields and will adjust their habits accordingly. A braconid, Microhracon cepJii, recently described by Mr. A. B. Gahan,' also attacks the larvse in grass stems, kills them before maturity, and spins a gray parchment-like cocoon within the gallery, generally near its lower end. This cocoon is truncate at both ends, its disklike extremities completely filling the bore. The adult escapes by biting an opening through the stem in the vicinity of the cocoon. ARTIFICIAL CONTROL From the foregoing sketch of the life history of the western grass- stem sawfly it seems obvious that this pest will have to be attacked while it is in the larval state. The egg and adult stages are both brief and are clearly beyond the reach of control measures of any sort. For nearly 11 months the insect exists as a helpless larva, protected only by the grass or grain stem within which it lives. If this stem could be destroyed, the larva witliin would perish. The first remedy that occurs to the farmer or the student of field conditions is the burning of the stubble in the autumn or spring. It would seem a very simple matter to set fire to the stubble and destroy at least the majority of the sawfly larvae that are hibernating in it. But when one begins to examine the infested fields it is found that the inhabited stems have been cut at the ground level or below so that it is often necessary to brush away the earth in order to find the stubs, containing the larvse. So little heat is generated when stubble is burned that these subterranean stems could not possibly be harmed by the quick passage of the flames. In 1907 Mr. Norman Criddle, in Manitoba, wishing to make a thorough test of this particular remedy, spread a layer of straw several inches deep over an infested area in a wheat field and set the straw on fire. More heat was produced than stubble alone could possibly make, the surface of the ground being too warm for the hand after the fire had died down. Even after this severe treatment it was found that, as far as could be learned by a minute search, not a single larva had suffered. They had simply retreated to the lower end of the hibernation cell and "kept cool." Another fact must be noted in this connection. When a field has been damaged seriously by the sawfly, the stubble remaining to feed a running fire is of necessity more scanty than in an uninjured field ' Gahan, A. B. Description of a new hymenopterous parasite (Braconidn^). In Proc. Ent. See. Wash.., V. 20, no. 1, p. 18-19. Jan., 1918. THE WESTERN GRASS-STEM SAWFLY. 25 and consequently it would be exceedingly difficult to burn such a field even under the most favorable conditions. In Utah the bunch grass, Elijinus condensatus, is much infested by this same fly and frequently is burned by fires that sweep the mountain side. This Elymus forms dense sods, with stems often more than 3 feet in length, and the heat from its combustion is great. The writer has examined a large series of burned sods and has seldom discovered any injury to the larvae from the fire. These facts would indicate the futility of burning the stubble as a control measure. Although it might seem possible to decrease the numbers of the fly by mowing roadside and fence-row grasses during July, thus destroying the larvie always present in the stems of these grasses, careful study has proved that a large percentage of the larvae in these grasses is parasitized and therefore it would seem unwise to take steps that might diminish the number of parasites. Without any doubt grain fields in North Dakota and Canada are invaded regularly by sawflies that issue from grass growing along their borders. Still, because of the multiplication of useful parasites from tliis same grass it is probably inadvisable to mow the grass in midsummer. Deep plowing, 5 to 6 inches, is perhaps the best remedy for the sawfly that can be suggested at present. It is much easier to advise tliis than to put it in practice. In almost every plowed field in any part of the country each furrow is marked by a row of stubble projecting from the inner edge of the furrow slice. Unless the stubble is turned squarely upside down, burying it at least 5 inches, the resulting surface at the same time being compacted by harrowing or rolling, the flies will be able to escape with ease from beneath the ground. In the fall of 1916 the writer buried four lots of infested stubble in different depths of earth sifted and compacted by jarring. These were buried, one at 3 inches, one at 4, and 2 at 6 inches, in glass jars, 10 stubs in each of the first two, 20 in the other two. August 6, 1917, these cages were examined with results as follows: Under 3 inches of earth all adults emerged. Under 4 inches 1 larva died, all others emerged. Under 6 inches 1 adult died in the cell, 6 larvae also died, 2 active living larvae still in the cell, all other adults emerged. Under 6 inches 7 larvae died in the cell, all other adults emerged. Lumpy soil in the field might make it easier or harder for adults to emerge than fine soil in a jar, and this point might be difficult to determine. Cultural conditions in North Dakota are not favorable for burying the stubble by plowmg. Spring wheat is followed in many cases by winter rye which is disked into the wheat stubble after harvest. This 26 bulli<;tin 841, u. s. depaktmknt of agriculture. procedure leaves all the infested stems of wheat on the surface, and nothing could be more favorable for the escape of the adult flies in the. following spring. The wheat stubble seems to be necessary to hold the winter snow for the protection of the young rye, hence the farmers seldom or never plow the stubble under before sowing the rye. Previous to the year 1919 it had been stated with much confidence by men who were known to be good observers that durum wheat was nearly immune from the attacks of the sawfly. On the strength of these statements county agents were inclined to recommend a modi- fication of ordinary farm practice, at least to the extent of barring from that region Fife and Marquis and the softer-stemmed wheats in the hope that by this means the work of the sawfly might be checked and a more certain harvest assured. It was readily seen that an immune wheat would solve the problem of the sawfly. Observations made by the writer during the month of August, 1919, and recorded on an earlier page of this paper, included in their scope an inquiry into this question concerning the immunity of durum wheat. Farm work was too far along at the date of this visit to permit of effective field work to settle the matter definitely, but several farmers informed the writer that durum had suffered severely that year, although not as much as either Fife or Marquis. These reports must be accepted at their face value since the agree- ment on this point was general. The immunity of durum may vary from year to year and is possibly based on the relative dates of the appearance of the adult Cephus and the rapidity of growth of the young grain. The stem of the durum wheat is more dense and unyielding than that of other wheats, and if a warm rainy spring should hasten its growth it might prevent the sawfly from placing many eggs. A number of unknown factors enter into this problem that hinder its complete solution at the ]) resent time. CEPHUS PYGMAEUS L. In certain parts of the country the occurrence of Cephus cinctus appears to have t)ecn confused with that of its congener Cephus pyg- maeus, a sawfly stem-borer that was probably imported from Europe only a few years previous to the first mention of Cephus cinctus in the United States. The habits of the two species are so similar that a brief synopsis of the life history of Cephus lyygmaeus is given herewith together with a condensed description of the same insect. As far as is now known the imported species does not yet occur west of the Mississippi River, while the western grass-stem sawfly has been found for the most part only west of the same river. C. pygmaeus was first observed in 1887 in the vicinity of Ithaca, N. Y,, and in 1889 Prof. J. H. Comstock published ^ an account of its THE WESTERN GRASS-STEM SAWFLY. 27 life history as worked out by himself. From this bulletin the follow- ing summary of its habits and appearance has been compiled. The adults, in the latitude of Ithaca, appear during the month of May and at once begin oviposition in the stems of wheat, just as the grain is jointing. In the majority of instances the eggs are placed above the third joint, and the larval gallery extends from the point of oviposition, or a little above, to the extreme base of the plant. By the time the wheat is ready to cut, early in July, a large majority of the larvse have descended to a position below the level of the reaper cut and are safe from removal with the harvested straw. A week later nearly all the larvae have girdled the stems within and part have already spun the silken lining of the hibernation chamber. CepJi.us inigmaeus is a well-known species in Europe and has been described by both English and Continental writers. In France it has been considered a very serious pest and is said to attack both, wheat and rye. DESCRIPTION Adult shining black, banded and spotted with yellow. Length of male 8 mm., of female 10 mm. Head large with prominent eyes ; three ocelli near summit. Antennae inserted on front nearly opposite middle of compound eyes, and composed of 19 or 20 segments. Wings transparent, iridescent, somewhat smoky, with costal margin yel- low toward base. Mouth-parts (except tips of mandibles), a spot on clypeus, a nar- row margin between compound eyes and mouth-parts, ventral aspect of thorax, legs (except a dark band on coxae and femora), membrane at base of abdomen, caudal margin of each abdominal segment ventrally, a more or less well-marked spot on each side of first and second abdominal segments, a broad band occupying caudal three- fourths of third and fifth segments, a narrow band on caudal margin of sixth segment often more or less interrupted forming spots on back and sides, and latero-caudal angles of seventh segment, yellow in male. In female, spots and bands usually smaller and sometimes entirely wanting ven- trally. Both this species and Ceiilius crnchis vary greatly in their mark- ings. Mr. S. A. Eohwer, of the Bureau of Entomology, who has given much critical study to this genus and has examined a large series of individuals, mostly reared from known host plants, states in a recent publication: ^ "The introduced, European, Cephus pygmaeus (Lin- naeus) is very similar to the native species common throughout the west and it is difficult to find characters which distinguish the two in all their forms." 1 Comstock, J. H. A saw-fly borer in wheat. In Bui. 11, Agr. Exp. Sta. Cornell Univ., Nov., 1889. 2 Rohwer, S. A. The American species of the genus Cephus Latreille. In Proc. Ent. Soc. Wash., v. 19, p. 139-141. 1917. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCTJEED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 10 CENTS PER COPY UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 843 ^ i yL^'^'«-r<^ Contribution from tiie Bureau of Entomology L. O. HOWARD, Chief j)J^'<^WU Washington, D. C. PROFESSIONAL PAPER June 7, 1920 THE BEAN LADYBIRD. By F. H. Chittenden, Entomologist in Charge, and H. O. Marsh, Entomological Asftistant, Truck-Crop Insect Investigations. With a report on "The Bean Ladybird in Colorado, in 1919." By A. C. Mallory, Scientific Assistant. CONTENTS. Introduction i^ynonymy Dfscription The adult The egg The larva The pupa Distribution Danger of future spread Life history and habits Seasonal history Reproduction and development- History and literature Unpublished records Natural control Effect of climatic conditions Natural enemies Page. 1 2 2 3 4 4 5 5 7 7 7 10 11 13 13 14. Page. Preventive measures 15 Hand picking and brushing 15 Clean cultural methods 15 Early and late planting 15 Direct measures of control 16 Experiments with insecticides__ 16 Cooperation IS Summary of control measures- _ 18 General summary 18 Literature cited 20 The bean ladybird in Colorado in 1919 Life-history records Injury Control 21 21 23 INTRODUCTION. In the semiaricl region of the Southwest, where an immense acreage of beans is grown annually, a destructive insect known as the bean ladybird, bean beetle, " bean bug," and spotted bean beetle ^ does great injury. Indeed, it is to the bean crop what the Colorado potato beetle is to the potato, a pest of the highest importance in the region which it inhabits. The beetle devours all parts of the bean plant — leaves, flowers, and growing pods. Its main food supply, however, is the leaves, through which it cuts irregular holes (PI. I; III; IV, fig. 1). Its injuries, ^ Epilachna corrupta Muls. ; order Coleoptera, family Coccinellidae. Note. — This insect was under the observation of the junior author (who died Sep- tember 10, 1918), from 1914 to 1917. The life-history investigations were conducted in an open-air insectary at Rocky Ford, Colo. ; and .some field operations were conducted also at Pueblo, Fort Collins, and Colorado Springs, Colo., and at Maxwell and French, N. Mex. 152117°— 20— Bull. 843 1 A 4»^ cv^^ g^ y^ 6 2 BULLETIN 843, U. S. DEPARTMENT OF AGRICULTURE. SO far as is known, are practically confined to beans, and no variety seems to be exempt from injurious attack. This insect has been ob- served feeding on various forms of the kidney bean {Phaseolus vul- gans et al.), including string, pole, navy, and teparj' or Mexican, and on the lima bean {Ph. lunatus). Of these string beans are favorites. On one occasion the soy bean {Soja hispida) was at- tacked. The beetles, unlike those of the bean leaf -beetle {Cerotoma trifurcata Forst.), show no tendency to injure very young plants and the larvae work on the lower surface of the leaves, skeletonizing large, irregular areas without cutting the epidermis or upper skin. (PI. n,iii.) It is fortunate that its field of operations is limited, both as re- gards the crop plants affected and the ten-itory over which it ranges. ■It has been estimated that it does an annual damage in New Mexico varying from 5 to 100 per cent of the crop, the average loss being conservatively placed at 10 per cent. This species is remarkable in that it is one of two species of lady- birds occurring in the United States^ which feed exclusively on vegetation, the other forms of the ladybird family being predacious and subsisting largely on plant-lice, or aphids, and the eggs of insects. SYNONYMY. The bean ladybird was described by Mulsant in 1850 (1, p. 815)- under the name by which it is known in economic literature, Epi- lacliiui corrupts. In the original description in which this name appears, j^rinted under Xo. 90, E. varivestis is also described as No. 91, yet Crotch (3, p. 62), followed by Gorham (8, p. 242), recognized the latter as the proper name for the species, and relegated coiTupta to synonymy, in which case the strict law of priority has not been follow^ed. This species has evidently been described under at least a half dozen names, but as there is no means of deciding positively the exact term to apply to the species under consideration, Epllachna corrupta is here used to avoid further confusion, although E. varipes Muls. was described first and is acknowledged by Crotch and Gorham to be the same species. The name used by Bland (2), E. maculiventris, described in 1864 from the Rocky Mountain region of Colorado, undoubtedly applies to this species and naturally falls into synonymy. DESCRIPTION. THE ADULT. The adult (fig. 1, b) is a robust beetle, oval in outline, and about one-fourth of an inch in length by about one-fifth of an inch in ^ The other is known as the squash hidybird (Epilachna horealis Fab.). - Reference is made by figures in parentheses to' " Literature cited," p. 20. THE BEAN LADYBIRD. 3 width. The color of the newly developed adult is yellow, gradu- ally darkening with age to a grayish brown. Each elytron or wing- cover is marked with eight small black spots of variable size. Technical descriptions of the genus and species follow : Genus Epilachna. Large, pubescent species related to Clulocorus. Sides of protlioras only slightly curved and broadly explanate ; those of. elytra rather strongly re- flexed ; epipleurge horizontal, broadly concave, not distinctly extended to sutural apex. Metasternal and ventral lines well-defined, legs moderately re- tractile ; femora not deeply sulcate beneath, tibiae with an acute external edge, and shallow groove for reception of tarsi ; claws cleft, with lower cusp nearly as long as upper. EPILACHNA COEEUPTA MTJLS. Form oblong, more narrowly oval than horealis and distinctly smaller, dull in luster, densely pubescent, and very closely, unequally punctate ; color gray- ish brown ; head and pronotum without spots. Each elytron ornamented with Fig. 1.- -Tlie bean ladybird (EpUachna corntpta) : Oj Larva; b, beetle; c, pupa; d, egg mass. About three times natural size. eight spots or dots of varying size in three rows; three small sub-basal spots in a broken row, median less basal ; three in a transverse subparallel row just before the middle, usually larger than sub-basal, median usually a little larger, and two near apical fourth, placed near inner fourth and outer third. Lower surface darker or concolorous with legs, which are pale throughout. Length 6.5-7.S mm. ; width 4.8-5.4 mm. The so-considered Mexican variety, E. varipes Muls., differs mainly from the species under discussion as it occurs north of Mexico in having the two subapical spots imited or coalescing, forming an arcuate fascia. All spots are also larger and surrounded by a lighter aureole. The typical varivestis, as figured by Gorham, shows these aureoles, but they are less pronounced in many specimens from the United States. Specimens occasionally occur of a more or less pronounced buff color, but these usually are not fully colored, being more or less im- mature when killed for mounting. THE EGG. The Qgg is dull pale yellow, elliptical in outline, approximately twice as long as wide, a little larger at the base or attached end than 4 BULLETIN 843, U. S, DEPARTMENT OF AGEICULTURE. at the apex. The surface is strongly sculptured. The length of the egg is 1.25 mm. ; the width 0.6 mm. The eggs are deposited on end on the lower surface of the bean leaves (PI. IV, fig. 1) in irregular masses, varying from half a dozen to about 75, with an average of between 40 and 50. (See fig. 1, d.) THE LARVA. The larva is j^ellow and the body is armed with long, formidable, strongly branched spines, darker at the tips. When full grown these spines, although somewhat irregular, are arranged in rows both longitudinally and transversely, as shown in the accompanying illus- tration (fig. 1, a). The head is moderately prominent, as are the mandibles and other mouth-parts, which with the eyes are darker than the other parts. The legs are rather long and stout and the anal segment is obtusely and roundedly produced. The newly hatched larva measures about one-sixteenth of an inch in length and the full- grown larva about five-sixteenths of an inch. Descriptions of the larval stages, as given by Merrill (21), follow: First Stage. When the yellowish larva first frees itself from the effs the spines are closely appressed. As the chitin dries, the spines become erect and are seen to be branched at and near the tip. Later the tips of the branches be- come darker. The larva is about 1.3 mm. long by .6 mm. wide. The body tapers sharijly in the abdominal region and is recurved downward. There is a row of four spines across the front of the rather pronounced pro-thorax. On the rest of the body there are six longitudinal dorso-lateral rows, the spines of the outside rows being very small and very few. * * * Second Stage. After the first moult the larva is 2 mm. long and the tip of the abdomen is slightly more curved than in the first stage. The spines are longer and more branched. The dark tips are not so pronounced. The rows of spines are the same in number but more distinct. * * * Third Stage. After the second moult the larva is 4 mm. long. The spines are longer, more branched from the sides, and dark tipped. The rows are now easily seen. In this stage the larva seems to be rather humpbacked, the highest and widest portion of the body being about the middle portion. The abdomen tapers sharply, the anterior end slightly, only. * * * Fourth Stage. At the beginning of the fourth stage the larva is 5.4 mm. long and it increases to nearly 1 cm. in length before the fourth moult. The chief difference between the larva in this stage and in the last is in the size. THE PUPA. The pupa (fig. 1, c; PL Y) is ovate in outline, and approximately the size of the adult. It is yellow with brown markings. Anteriorly it is roundedly subtruncate and posteriorly tapers strongly toward the apex. The surface is sjparsely beset with bristle-like setae and long hairs. The head is folded down in front over the thorax and the posterior legs reach below the wing-sheaths. The apex terminates in two elongated processes, conical at the base and black at the ex- treme tips. The larval exuviae are pushed down and form a protec- tion for the last abdominal third of the body. (PI. VI.) Length 7-7.5 mm. : width 4-4.5 mm. THE BEAN LADYBIRD. DISTRIBUTION. The bean ladybird occurs in the States of New Mexico, Arizona, Colorado, and Texas. In the last mentioned State it is restricted to localities in the western part. It is also said to occur in Kansas and Arkansas but not as a pest. There is a single record of a beetle, presumably a "chance find," having been taken at Ogden, Utah, but it is not positively known that the species breeds in that State. The bean ladybird, as has been stated, is to the bean industry in the West what the Colorado potato beetle is to the potato crop in the East, and its origin is obviously the same — Mexico, where it is widely distributed. It is also found throughout Central America. It has been observed at elevations of from 3,000 to 7,000 feet above sea level in New Mexico, and of about 4,000 to approximately 5,000 feet in Colorado. Fig. 2. — Map showing known distribution of tlie buan ladybird iu tlie United States, June 1, 1919. According to the observations of Dr. C. P. Gillette, State entomolo- gist of Colorado, the species inhabits the foothills of the Rocky Mountains, seldom straying far out on the plains, with the exception of down the Arkansas Valley in Colorado. Further details in regard to its distribution in that State are furnished on page 13. The known distribution of the bean ladybird in the United States is indicated in figure 2. DANGER OF FUTURE SPREAD. In regard to the future distribution of this pest it must again be compared with the Colorado potato beetle, first because of its ob- viously similar origin. It may have been introduced . at an even earlier date than the latter, since the cliif dwellers in the region which the beetle inhabits in the United States included tepary beans 6 BULLETIN" 843, U. S. DEPARTMEI^T OF AGRICULTURE. in their dietary. From an economic standpoint it is a potential pest of the type of the Colorado potato beetle and the boll weevil and it is singular that it has never migrated to any noticeable extent as have those pests, since it is probably capable of extended flights. One reason that may be assigned for this is its practical limitation to a. single food plant, while the potato-feeding insect infests virtually all of the Solanaceae, including the weeds, and is capable of breeding continuously on all species of Solanum and probably on other genera. There is no reason to suppose that this insect may not b}^ flight in- crease its present range materially some time, although not neces- sarily in the near future. Its further dissemination, however, would doubtless be slow and never as rapid as in the case of the compara- tively fleet-winged and more adaptable Colorado potato beetle. The reasons, then, for its failure to have become more widely distributed are : Its limitation to a single food plant and its probable incapacity for protracted flight with the wind. Moreover, it is probably not capable of inhabiting such varied climates as is the Colorado potato beetle, a species which seems to have no respect for life zones but which thrives equally well from subtropical southern Texas to boreal Manitoba. In the case of the Colorado potato beetle it can not as j^et be defi- nitely stated, as some authors have assumed, that it breeds wherever potatoes are grown, but it is perfectly capable of doing so, and it may be that in the course of time, many years undoubtedly, the bean lady- bird will be distributed wherever its food plant is cultivated.^ Another factor Avhich strengthens the belief that the bean ladybird will, in the course of time, become more widely disseminated is its very close relationship, both structurally and biologically, to the squash ladybird {Epilaehna horealis Fab.), which ranges from South America northward through Central America, Mexico, and the An- tilles, along the Mexican and Atlantic seaboard States to Maine and Canada. Obviously the squash-feeding species has a similar tropical origin, beginning farther southward and extending much farther northward. Instead of progressing straight northward it has fol- lowed more nearly the coastal lines and has a totally different distri- bution in the United States, being somewhat restricted to the East ^ just as the bean ladybird is restricted to the Middle West. The present distribution of this species as outlined in the map would indicate that we may expect its establishment some time in the future in near-by counties in the States of Utah, Wyoming, Ne- 1 The predacious ladybird Hippodamia convergens Gudr. is capable of accommodat- ing itself to practically all climes and countries, with the exception of areas where the temperature is so high or so low that few forms of plant and insect life are able to survive. - It occurs, though not as a pest, in certain other regions remote from the region speci- fied, e. g., in Kansas and Arkansas. THE BEAN LADYBIRD. 7 braska, and Oklahoma, and later in southern California. There may- even exist a wider distribution than is now known in Texas, since the localities inserted on the map plainly show such a possibility. LIFE HISTORY AND HABITS. SEASONAL HISTORY. In the Arkansas Valley of Colorado, and in regions having a similar climate, two generations or "broods" of the bean ladybird develop annually. The winter is passed in the adult stage, the beetles hibernating under tufts of grass, weeds, old vines, rubbish, and similar material, in or about the fields and gardens in which they developed. The overwintered beetles emerge from their hibernating quarters about the middle of June and, after a brief interval of feeding, mate and begin to deposit eggs. ■ The first eggs hatch in about a week and the adults of the first generation develop shortly after the middle of July. After an in- terval of a week or ten days eggs are deposited by the first genera- tion of beetles and from these the first adults of the second genera- tion develop. This occurs during the latter part of August or early September. A portion of the adults of the first generation and all those of the second generation deposit no eggs until June .of the fol- lowing year. The beetles go into hibernation during the last days of September and the first of October and, as previously stated, remain dormant until about the middle of June of the succeeding year. It is somewhat remarkable that the beetles remain in hibernation during the last days of May and the first half of June when high temperatures, from 90° to 95° F., often prevail. The egg-laying period of the overwintered beetles, which includes individuals of both the first and second generations, extends from shortly after the middle of June until about the 1st of August, although occasionally some of these beetles live and deposit eggs throughout the summer. The egg-laying period of the beetles of the first generation which deposit eggs during the first season extends from soon after the middle of July until well into September. Reproduction, then, continues from about the middle of June until the beans are destroyed by killing frosts in late September or early October. The insects usually cause a maximum amount of damage during July and August. The larvae, especially those more than half grown, are voracious feeders a;nd, as a rule, cause vastly more injury than do the beetles. REPRODUCTION AND DEVELOPMENT. The life-history studies of the bean ladybird were conducted in Colorado in an open-air insectary at Rocky Ford. The insects were confined in cloth-covered battery jars and fed on the foliage of 8 BULLETIN 843^ U. S. DEPARTMENT OF AGRICULTURE. string beans. A pair of beetles which developed in late August, 1914, mated August 29 and were isolated. These beetles commenced hibernation October 12, and on November 10 the rearing cage, con- taining earth and some dead bean leaves under which the beetles rested, was placed in the laboratory cellar. The cage remained in the laboratory cellar until May 20, 1915, when it was again placed in the open air. The beetles emerged from hibernation on June 15, and began feeding. The first eggs were deposited June 18, and from this stock the species was reared for two seasons (1915 and 1916) without a break. The record for 1915 is given in Table I and that for 1916 in Table II. Table 1.— Generations of Epilachna corruj)ta in 1915. Life-history event. First generation issued. Second generation appeared. Adults developed First eggs deposited. . . First eggs hatched First larvae matured. . First larvpe pupated. . . First adults developed Egg period Larval period Pupal period Total duration. . Aug., 1914 June 18,1915 June 2.5,191.1 Julv 12,191.5 July 1.3,1915 Julv 19,1915 Julv 19,1915 Julv 30,1915 Aug. 5,1915 Aug. 23,1915 Aug. 24,1915 Sept. 1,1915 Days. Days Table II. — Record of the generations of EpilacJina comipta in 1916. Lifc-historv event. First generation issued. Second generation appeared. Adults developed First eggs deposited. . . First eggs hatched First larvfe matiu-ed. - Firi^t larva^ pupated. . . First adults developed Egg period Larval period Pupal period Total duration. . Sept. 1,1915 June IS, 1916 June 25,1916 Julv 10,1916 .Julv 11,1916 Julv 17,1916 Julv 17,1916 Julv 30,1616 Aug. 5, 1916 Aug. 21,1916 Aug. 23,1916 Aug. 31,1916 Days. Days. 32 The 19 beetles which developed September 1 fed until September 23, when hibernation began. The cage was placed in the laboratory cellar November 10, 1915, and was removed to the open May 3, 1916. June 12, 16 beetles issued from hibernation, 3 having died during the winter, and began feeding. One pair mated June 13, and the record of the progeny is given in Table II. The female died September 9. The male went into hibernation October 5, and was later destroyed. In this case the egg-laying Bui. 843, U. S. Dept. of Agriculture. Plate I. The Bean Ladybird. Beetles beginning work on under surface of leaf. Bui. 843, U. S. Dept. of Asriculture. Plate 1 1. The Bean Ladybird. Bean leaves showing injury by beetle. Bui. 843, U. S. Dept. of Agriculture. PLATE III. The Bean Ladybird. Bean leaves skeletonized by the bean ladybird. Beetle above; pupae in middle; larva at left near bottom. Somewhat enlarged. Bui. 843, U. S. Dept. of Agriculture, Plate IV. Fig. I.— Egg Mass on Lower Portion of Bean Leaf. Fig. 2.— Larv/e at Work on Under Surface of Leaf. THE BEAN LADYBIRD. Bui. 843, U. S. Dept. of Agriculture. Plate V. The Bean Ladybird. Pupee formed on injured bean leaf. Bui. 843, U. S. Dept. of Agriculture. PLATE VI. The Bean Ladybird. Cluster of bean leaves showing natural form of pupation on lower surface of leaf, and larva ready- to transform on leaf at right. THE BEAN" LADYBIRD. 9 period extended from July 25 until September 2, a total of 40 days. With other females this period covered from 35 to 70 days. The life cycle from egg to adult and the periods of the stages by days are given in Tables III and IV respectively. Table III. — Life cycle of Epilachiia rornipta. Event. Eggs deposited. Larvae hatched Larvae molted. Larvae molted. Date. 1916. Aug. 10 Aug. 18 Aug. 21 Aug. 25 Event. Larvae molted Larvae reached maturity Larvae pupated Adults developed Date. 1916. Aug. 30 Sept. 3 Sept. 5 Sept. 12 Table IY. — Periods of stafies of Epilachna corrupta. Instar. Egg period First larval stage period. . Second larval stage period Third larval stage period. Number of days. Instar. Fourth larval stage period I'upal period Total Number of days. The beetles which developed August 31 fed, without mating or depositing eggs, until October 5. when they began hibernation. No- vember 10. 1916, the cage was placed in the laboratory cellar and removed to the open April 17, 1917. The beetles issued from hiberna- tion June 15, 1917, and were then killed, closing the record. Records of egg-laying Avere obtained by confining single pairs of beetles, immediately after mating and before the first eggs were deposited, in small rearing jars. The beetles fed on bean leaves and the eggs were removed as deposited and counted. The number of eggs deposited by eight females of the first and second generations were: 504, 552, 616, 636, 850, 942, 1,147, and 1,516, respectively, or an average of 845 eggs to a single beetle. A detailed record of the female of one of these pairs of beetles, which developed July 17, 1916. and mated July 22, is given in Table V. Table Y. — Egg-laying record of a single female of Epilaehna corrupta in 1916. Date. 1916 July 25 July 28 July 29 Aug. 1 Aug. 3 Aug. 5 Aug. 6 Aug. 9 Aug. 11 Number of eggs deposited. Date. 1916 Aug. 15 Aug. 18 Aug. 21 Aug. 23 Aug. 26 Aug. 28 Sept. 2 Total Number of eggs deposited. 56 58 55 52 52 54 51 850 152117°— 20— Bull. 843- 10 BULLETIN 843^ U. S. DEPARTMENT OF AGRICULTURE. In experiments conducted in New Mexico by Merrill (21) the period of incubation was between 4 and 9 days. The duration of the larval stage was between 15 and 21 days. The pupal period was between 3 and 5 days, and the total developmental period lasted be- tween 22 and 28 days. HISTORY AND LITERATURE. The bean ladybird was described by Mulsant in 1850 (1) from Mexico as EpUachna corrupta and its injuries were observed in New Mexico at about the same period. In 1883 Riley (4) published an editorial on this species witli quo^ tations from a letter from Prof. George H, Stone, which contains our first known account of the food and injurious habits of this insect. Attack by larvae and adults was observed on leaves and pods of wax beans at Colorado Springs, Colo., August 26, 1882. It was not until a lapse of six years that attention was called to further injury by this insect. At that time Judge J. F. Weilandy (5, 6) wrote, July 23, 1889, of injuries at Springer, N. Mex., stating that this " bean bug " committed great depredations in bean fields, often destroying them entirely. The Mexicans had found that late planting, about the middle of July, acted as a preventive of its rav- ages. In a letter dated July 30, in the same year, he directed atten- tion to injury at Watrous, N. Mex., and stated that the pest had been known in that region for about 40 years. He also furnished speci- mens from which were recorded, editorially, the habits of two pre- dacious ladybirds, which will be mentioned later under " Natural enemies," of feeding on the eggs of this species. In 1892 Prof. C. P. Gillette (7) gave an account of this pest in Colorado, furnishing illustrations of the stages and manner of work. In 1897 Eev. Henry S. Gorham (8) considered this species with the Coccinellidae of Central America, indicating the synonymy, and the distribution in Mexico, Guatemala, and Panama, with notes on variation and colored illustrations of the adult and of the larva. In this year also Mr. H. Griffin (9) reported injury at San Juan, N. Mex. Judge Weilandy furnished a list of the varieties of beans affected and reported on the effectiveness of Paris green which, although it killed the beetles, destroyed the plants as well. In 1899 Col. Thos. L. Casey (11, p. 103) furnished a characteriza- tion of the beetle, comparing it with E. horealis. In 1900 Cockerell (13) stated that this species was the " bane of bean growers in New Mexico, from Chicorico Canon * * * to the Mesilla valley." In 1902 A. N. Caudell (14) cited an instance of extreme injury to beans at Fort Collins, Colo., in 1901. The statement made by W. Knaus (15), in 1906, that this insect was damaging potato near Wootens, N. Mex., is, of course, incorrect. During 1907 Messrs. THE BE AX LADYBIRD. 11 Fall and Cockerell (16, p. 170) indicated by localities the distribu- tion in New Mexico. In 1913 Dr. A. AV. Morrill (19) published a note on the distribution of this species in Arizona. In 1915 E. O. Essig (20, p. 219) stated that this species was "said to be found in California," which is evidently incorrect as no definite locality is cited and the species is not known to breed in that State. During 1917 D. E. Merrill (21) published the first comprehensive account of this insect, with especial regard to its occurrence in New Mexico, furnishing details in regard to injuries, life economy, and distribu- tion, and indicating methods of control. In most respects the results obtained in that State do not differ materially from those obtained in Colorado. Popular accounts of the bean ladybird were published by Gillette (10) in 1898, Sanderson (18) in 1912. and by the writer in 1899 (12), 1907 (17, p. 109), 1917 (22, p. 28), and 1919 (23), as well as by others. UNPUBLISHED RECORDS. February 2, 1899, Mr. C. B. Metcalfe wrote of the bean ladybird and its injuries to the Mexican Bayo bean or frijole at San Angelo, Tom Green County, Tex. For many years prior to the date of writing growers had not succeeded in raising a crop of beans because of the ravages of this pest, which destroyed the plants by eating the leaves. Metcalfe described the larva as a humpbacked yellow insect about one- fourth inch long and of the color of sulphur, with a hairy- looking covering which changed afterward to the hard-backed grown " bug." He described the larva as destroying the green part of the leaves leaving only a thin tissue. During the same year Mr. James K. Metcalfe, Silver City, N. Mex,, wrote that this species was quite injurious to beans in his vicinity, and furnished specimens in different stages. At this time, September 14, larvae were quite scarce, most of those sent having transformed to pupse. August 8, 1904, Oscar Liffreing, Bernardo, N. Mex., sent specimens, mostly pupae and newly developed beetles, with the report that they were devouring all of the early beans in that region. August 26 of the same year specimens were received from Mr. Liffreing, with re- port of injury to beans. July 26, 1905, Mrs. V. A. Armstrong re- ported injury to beans at Fort Collins, Colo., furnishing specimens of larvae, and leaves and pods showing injury. September 6, 1908, F. H. Headley reported injury at Fort Collins, stating that this insect was doing a great amount of damage to the bean crop in that section. July 15, 1909, B. F. Morris, Santa Cruz, N. Mex., wrote as follows : " I am sending some chinch bugs, the in- sect which is working destruction to the bean crop here, and is now depositing its eggs." During 1909 M. C. Stevenson, Espanola, 12 BULLETIN 843, U. S. DEPARTMENT OF AGRICULTURE. N. Mex.. reported injury, July 24, to beans, furnishing samples of adults. August 11, M. A. Bishop reported injury at Tularosa, N. Mex. He wrote that for two years previously a small yellow bug covered with hairs had been eating the leaves of beans leaving noth- ing but a bare skeleton. Complaint was also made by Mrs. Katherine Courtney of injury to lima beans at Littleton, Colo. May 28, 1910, Elias Clark, Alcalde, N. Mex., stated that this insect was very destructive to beans in that vicinity. January 11, 1909, John Block, Santa Cruz, N. Mex., described the larvae, adults, and the work of this species on beans, requesting a remedy. July 6, 1915, Mrs. Ethel Mercer wrote from Denver, Colo., of little yellow larvse which literally devoured the bean crop. They began on the leaves, and after these were gone they attacked the pods, as many as 15 being noted on a pod. Before the beans began to bloom the same insect ate the leaves " full of holes.'' August 28, 1912, this species, according to Dr. A. W. Morrill, did considerable damage to beans at Prescott, Ariz. At that time it was abundant mostly in the pupal and adult forms. Between July and August of the following year, 1913, much damage was noted to pole beans at Cottonwood, Ariz. The insects practically ruined an entire field of about one-half acre. At that time full-grown larvae and pupae and a few egg batches and young larvae were found. The hibernating adults had disappeared, only newly emerged, pale yellow adults being in evidence. October 13, 1916, W. E. Marble wrote in regard to the growing of Mexican beans in the Arroyo Animas Valley of New Mexico, stat- ing that the crop was greatly damaged by the larvae of this lady- bird which he described as of about the size of a navy bean, yellow in color, eating the leaves, and leaving onlj^ a network. This, he wrote, stops the growth of the plants and ultimately kills them. October 16, A. Warner, Sandy, Ariz., sent specimens, stating that the insect destroyed a crop in about 14 days, and that Paris green was ineffective. Complaints were made during the year 1917 of injury at IVlieat- ridge, Brewster, Colorado Springs, Rocky Ford, Pueblo, Denver, and Boulder, Colo. ; Santa Fe and East Las Vegas, N. Mex. ; Flag- staff, Ariz. ; and Alpine, Tex. The last report was from T. F. Blaine, dated October 31, and was accompanied by specimens. D. E. Merrill also wrote of this species October 5, 1917, and of its oc- currence in the vicinity of El Paso, Fabens, and Clint, in the El Paso Valley of Texas. F. M. Wadley, scientific assistant, stated June 21, 1918, that the bean ladybird occurred at Wichita, Kans., but was more abundant in the western part of the State. August 6, Thomas H. Hudson complained of injury by this species to beans at Colorado Springs, Colo. THE BEAN LADYBIRD. 13 During January, 1919, A. E. Mallory, scientific assistant, wrote in regard to the occurrence of this species at Greeley, Colo., that he had observed the beetles feeding on the underside of the leaves of the mammoth soy bean {Soja Jiispida) in that vicinity. The previous December he found the beetles hibernating under a pile of bean vines. Until the receipt of this last record the species was believed to feed exclusively on table beans. February 11, 1919, W. A. Williams, Venus, N, Mex., reported in- juries by this species to pinto beans in that vicinity. Failure to obtain rain was ascribed as one of the causes of injury. Mr. Wil- liams stated that " on beans these little ' bugs ' eat the leaves full of holes and damage the crop considerably." During January and February a number of other complaints were made of injury by this species in New Mexico and Colorado. Other complaints of injury, unaccompanied by specimens, and requests for information in regard to methods of control have been received from other sources, notably Parsons, Kans., Jemez Springs and Nogal, N. Mex., Griffith, Colo., and Kirkland, Ariz. April 3, 1919, the State entomologist. Dr. C. P. Gillette, collabora- tor, wrote in regard to the occurrence of this species in Colorado as follows : We have records of this species from tlie points which you mention and others near them, and also from Nucla, in the Paradox Valley, Montrose County ; from Cortez, in Montezuma County ; and from r.ear Greeley, in Weld County. It seems likely that the infestations in the southwestern part of the State are from New Mexico. It has seemed very strange to me that this beetle keeps so close to the foothills, never going out far upon the plains, except down the Arkansas Valley. The beetle was abundant here when I first arrived at Fort Collins, 28 years ago, and it apparently has never occurred as far east as Greeley, about 24 miles from the foothills. I have found it to be a species about equally abundant every year, although there is some fluctuation in num- bers. It was very bad last year in the northern section, from Pueblo to Fort Collins, along the eastern foothills, and extended a few miles into the plains. Writing on this species September 2, 1919, Mr. Fabian Garcia, hor- ticulturist. New Mexico Agricultural Experiment Station, State Col- lege, N. Mex., stated that this insect is a serious pest in New Mexico, particularly in the older bean-growing sections, and that it causes the bean growers there a lot of trouble and expense. He expressed the opinion that the losses could be materially reduced by properly spraying. NATURAL CONTROL. EFFECT OF CLIMATIC CONDITIONS. Cold weather appears to be the most important natural check to the development of the bean ladybird in Colorado. This insect is a southern species which apparently has not become fully adapted to 14 BULLETIN 843, U. S. DEPARTMENT OF AGRICULTURE. 1 Fig. 3. — The five-spotted ladybird (H ippodamia ^-signata) , an enemy of the hean ladybird. Enlarged. northern climates. This is not a theory but is based on facts. It is indicated by the fact that many eggs, larvae, and pupae occur so late in the fall that they are killed by freezing. Many larvae starve be- cause the foliage of the bean plants on which they had been feeding was destroyed by early fall frosts. It is also not uncommon to find many dead adults in their hibernating quar- ters where they had been killed by winter temperatures. NATURAL ENEMIES. The insect enemies of the bean lady- bird are, as far as has been learned, not particularly effective in holding it in abeyance. The beetles are well pro- tected by their firm elytra or wing cov- ers and by a repellent yellow liquid which oozes from their knee joints in small drops when the insects are disturbed. This liquid possesses a disagreeable odor and doubtless a similar flavor, which, it is believed, may protect the beetles from the attacks of natural enemies. June 27, 1916, two overwintered female beetles were collected at Rocky Ford, Colo., each with the egg of a tachinid fly attached to one of its elytra. One of these beetles died September 4, and the other Sep- tember 9. Fertile eggs were depos- ited at frequent intervals from June 29 until September 5, and no parasites developed. This is the only evidence noted of insect parasites. Morrill has reported an undeter- mined ant observed eating the eggs on one occasion. The adults of three species of pre- dacious ladybirds are known to be natural enemies of this insect. They are: The convergent ladybird {Hippodcmiia convergens De G.) ; the five-spotted ladybird {Ilippo- damia B-signata Kby., fig. 3), and the transverse-spotted ladybird {Coccinella transversoguttata Fab., fig. 4). These have been reported as destroying the eggs of the bean lady- bird, and next to cold are the most effective known factors in its natural control in Colorado and neighboring States. Fig. 4. — Transverse-sixitted lady- bird (Coccinella transversogut- tata) : Adult beotle. Much en- larged. THE BEAN LADYBIRD. 15 The first species occurs in abundance throughout the country and is our most useful ladybird, having been transported from one part of the country to another and to foreign countries. The other two are commonly found in the region inhabited by the bean ladybird, but more especially in the middle Northwest. Both species, how- ever, extend their range to Washington and Oregon. The larvae are apparently well protected from insect enemies by the branched spines with which the body is armed. In one case, however, the larva of a lacewing fly {Chrysopa sp.) was observed sucking the juices from a partially grown Epilachna larva. No insect enemies of the pupa and no fungous or other disease have been observed to affect the living insect in any stage. PREVENTIVE MEASURES. HAND PICKING AND BRUSHING. The bean ladybird is difficult to control. In small gardens hand- picking the eggs, larva\ and adults has given satisfactory results. The greatest measure of success has come from gathering and destroy- ing the overwintered beetles soon after their emergence from hiberna- tion and before they have had an opportunity to deposit eggs. The beetles, being sluggish like the Colorado potato beetle, are readily hand picked. Another method wdiich has afforded some degree of success consists in brushing the larvae from the foliage to the earth between the rows. This can be accomplished by striking the plants with the bare hand, with a bunch of weeds, or with a paddle fashioned for the purpose from a shingle. If the brushing is done during dry hot weather very few, if any, of the larvae are able to return to the plants. CLEAN CULTURAL METHODS. With the knowledge that the adults of the bean ladybird pass the winter under old vines, tufts of grass, weeds, and other useless mate- rial, the numbers of beetles may be materially reduced by burning in late fall or early spring all rubbish of this nature along ditches and fence corners and in similar locations. Everything possible should be done to destroy these winter quarters, as their destruction will afford a considerable measure of protection from injury, if done by a community year after year. EARLY AND LATE PLANTING. Proper attention to the time of planting will prevent considerable injury by this as well as many other species of insect pests. By planting earlier than usual this can be accomplished, as well as by ]3lanting considerably later, or as late as a crop can be assured. Since the overwintered beetles do not begin to feed until very late, planting early will accomplish much, enabling the plants to make such good growth that insect damage coming late may be immaterial. Late planting should be so timed that the plants will come up after the overwintered beetles have about ceased feeding and, at the same 16 BULLETI:N' 843^ V. S. DEPAETMENT OF AGKICULTURE. time, early enough to secure a good crop before frost time. No definite time can be assigned for early or late planting for the entire range of this species; it is a matter for the growers themselves to determine. Community work should accomplish much along this line, in determining both the times for planting and the effect of this method. It has been suggested that early planting be practiced in a com- munity for a series of years and then late planting for a year or two succeeding this. Whatever can be done toward lessening the number of insects in a community during a given year will have a correspondingly greater effect for the coming season. DIRECT MEASURES OF CONTROL. EXPERIMENTS WITH INSECTICIDES. Numerous spraying experiments were made with arsenate of lead, Paris green, arsenite of zinc, and nicotine sulphate. The experi- ments were conducted on moderately infested plots of string beans. The spray was applied to both the upper and lower surfaces of the leaves with a portable compressed-air sprayer, fitted with an exten- sion rod, elbow, and disk-type nozzle having a fine aperture. ARSENATE OF LEAD. Arsenate of lead was applied at the rate of 1^, 2, 2 J, and 3 pounds in powdered form in 50 gallons of water. The spray adhered well and evenly to the foliage, but the effect Avas very uncertain on the bean foliage as was also the killing effect on the insects. The injury from burning varied greatly with the age and tenderness of the plants, the older, tougher foliage usually escaping appreciable in- jury, while on the younger, more tender plants the burning effect was serious, especially where the stronger doses were applied. In summing up the experiments in spraying with powdered arse- nate of lead the results were so uncertain that one is hardly justified in recommending this insecticide as a reliable agent for controlling the bean ladybird on string beans. One experiment was made with arsenate of lead paste at the rate of 6 pounds to 50 gallons of water. This burned the beans so badly that they were almost completely destroyed. Most of the larvse were killed, but the majority of the beetles escaped injury. This test indicates that paste arsenate of lead is even more in- jurious to bean foliage than the powdered form, and that the killing effect on the adults of the bean ladj'bird is equally uncertain.^ In experiments conducted by Merrill (21) in New INIexico pow- dered arsenate of lead was used at the rate of 2 and 2 J- pounds to 50 gallons of water without damage to the plants. Most of the beetles • Tbe tests herein mentioned were conducted with standard or acid lead arsenate. Neutral (diplumbic or triplumbic) load arsenate, in experiments at Washington, D. C.^ applied at standard dosage, caused no injury to bean foliage. THE BEAN- LADYBIRD. 17 left these plants at once. Larvje hatching from eggs deposited before the application of the poison were also killed where they fed on sprayed leaves. The older larva?, however, appeared to die of star- vation rather than from eating the poisoned foliage.^ ARSENITE OF ZINC. Experiments were made witli powdered arsenite of zinc at the rate of 1 pound to 20, 30, 40, and 60 gallons of water, respectively. The burning effect on the foliage was in all cases less than where arsenate of lead or Paris green was applied, and usually a larger proportion of the insects was killed. The burning was most ap- parent about the margins of the holes made in the leaves by the in- sects in feeding. Many larva? died after eating the poisoned foli- age, but, as with the other arsenicals, the effect on the beetles was uncertain. All factors considered, the most promising results were obtained with zinc arsenite at a strength of 1 to 40. This caused comparatively slight burning and killed an appreciable number of the insects. It should not be overlooked, however, that the killing of the beetles is uncertain and that the burning effect on the plants will vary greatly with their age and tenderness. It is probable that Mexican beans, with relatively tougher foliage, would show less injury from burning than the more tender-leaved string beans treated in these experiments. Zinc arsenite in experiments con- ducted in New Mexico by Merrill at the rate of 2 pounds to 50 gallons of water produced practically the same results as powdered lead arsenate. PARIS GREEN. Experiments M^ere made with Paris green at the rate of 1 pound to 60 and to 80 gallons of water. As a result of these tests the beans were destroyed by burning due to the presence of free arsenic. Most of the larva? were killed, but many of the beetles escaped. In the face of these results, Paris green at these strengths can not be recommended as a means of controlling the bean ladybird. Every- one who has tried Paris green has experienced the same failure. Sodium arsenite is at least equally dangerous. SUMMARY OF SPRAYING EXPERIMENTS. The experience of entomologists in spraying with arsenicals in Colorado and New Mexico tends to show that arsenate of lead acts largely as a repellent rather than as an insecticide, which is true also of its effectiveness in the case of such other pests as the striped cucumber beetle. Bordeaux mixture, which has come to be considered a standard re- pellent against flea-beetles, should be tested against the bean lady- bird in the future. ^ Powdered lead arsenate at the rate of 5 pounds to 50 gallons of water was no more effective but did no damage to the plants in experiments made. It should not be used at this strength. 18 BULLETIN 843, U. S. DEPARTMENT OF AGRICULTUEE. Additional experiments are necessary with arsenate of lead and arsenite of lime, alone and in combination with Bordeaux mixture. NICOTINE SULPHATE INEFFECTIVE. In another series of experiments larvae about one-fourth grown were sprayed with nicotine sulphate at the rate of 1 ounce to 2, 4, and 6 gallons of water, respectively. The larvae apparently were pro- tected by their spines and the applications were in all cases abso- lutely ineffective. The only noticeable effect was that the larvae ap- peared somewhat stupefied for a brief interval. COOPERATION. In the control of this pest, as with so many others which are dif- ficult to destroy, combined effort on the part of the bean growers of the community is essential to success. Whatever can be done in cooperation to lessen the numbers of this insect in one season is felt the next season and if it were rigidly continued would mean the eventual saving of the crop. SUMMARY OF CONTROL MEASURES. In the light of our present knowledge the best methods of con- trolling the bean ladybird may be summarized as follows : (1) For small gardens and similar areas hand pick the over- wintered beetles as soon as possible after they emerge from hiberna- tion. (2) Brush the larvae, or young, from the plants during hot, dry weather. (3) Spray with arsenite of zinc, at the rate of 1 pound to 40 or 1^ pounds to 50 gallons of water, or with arsenate of lead 1 or 2 pounds (powder) to 50 gallons of water. (4) Clean up the fields by removing dead grasses, weeds, and other possible hibernating quarters during the fall or winter months, and destroy them by burning, or by simply burning over the fields when this practice can be safely followed. (5) Early and late planting should be practiced. No specific time can be indicated for this that would apply to the entire region which the insect inhabits, and it is more satisfactory for the grower's to work out this problem for themselves. (6) In the case of large areas of beans, close inspection is strongly recommended. Infestation usually begins in small, localized areas, and if these infested spots are located and prompt measures, as indi- cated in the preceding paragraphs, are taken to destroy the insects a general infestation can be prevented, GENERAL SUMMARY. The bean ladybird is a serious pest on beans of all kinds, includ- ing the soy bean, in Colorado, New Mexico, Arizona, and western Texas. It frequently destroys entire crops and the conservative es- timate of the annual losses incurred is placed at 10 per cent of the crop. This insect feeds normally on the leaves, and attacks also the young pods and occasionally eats into the blossoms. • The beetles feed THE BEAlsr LADYBIRD. 19 chiefly on the upper surface, cutting irreguhu- holei5 in and through the leaves, while the larvae feed on the lower surface and skeletonize the leaves, seldom cutting through them. The adult is a robust beetle, about one-fourth inch in length, of oval outline, pale brown, with each wing-cover marked with eight small black spots. The larva is light yellow and anued with branched spines. The yellow eggs are deposited from about the middle of June until August on the lower surface of the leaves, in clusters of 40 or more, and sometimes to the number of 1,500 by a single female. The larvse feed at first in colonies, but with larger growth scatter and become more or less solitary. The life cycle may be passed in summer in from 22 to 30 days; the eggs hatch in from 4 to 9 days ; the larval period is between 15 and 21 days; and the pupal period varies from 3 to G days. In colder weather, however, these periods are longer. Two genera- tions or " broods " are produced annually. Cold weather in late autumn has the effect of destroying the in- sects, and their eggs are also destroyed by three species of pre- dacious ladybirds. In the control of the bean ladybird preventive measures are the most efficient, consisting of hand picking and brushing from the plants, clean culture, and early and late planting. Arsenicals possess some killing properties, but in the main act as repellents. Spraying with arsenate of lead, 1 or 2 pounds (powder), and witli arsenite of zinc, 1 to 1^ pounds (dry) , to 50 gallons of water, are the most prom- ising. Bordeaux mixture, 1 4 50 formula, should be used alone and in combination with these arsenicals. Additional tests must be made with these substances to ascertain the most effective and economical combination that may be applied to the bean plants without scorching or burning the leaves. For the treatment of large areas infested by the bean ladybird, a traction sprayer with nozzles arranged for side spraying of the type shown in figure 5 is useful. Owing to the danger of scorching bean foliage, it is desirable that an up-to-date sprayer should alwaj^s be used. Fig. 5. — Traction sprayer with nozzle arrangement for side spraying, of type useful for spraying beans for the bean ladybird. 20 BULLETIN" 843, U. S. DEPARTMENT OF AGRICULTURE. LITERATURE CITED. (1) MULSANT, M. E. 1S.50. Species des Col^opteres trimeres securipalpes. Paris. (2) Bland, H. J. 1864. Descriptions of new North American Coleoptera. In Proc. Ent. See. Phila., V. 3, p. 253-256. (3) Crotch. G. R. 1874. A revision of the coleopterous family Coccinellidse. London. <4) Riley, C. V. 1883. Epilachnn corrupta as an injurious insect. In General Notes, Amer. Nat., v. 17, p. 198-199. February. (5) WiELANDY, J. F. 1889-90. Injurious insects in New Mexico. In U. S. Dept. Agr., Insect • Life, V. 2, p. 113-115. (6) 1891. The New Mexican Epilachna. In U. S. Dept. Agr.. Insect Life, V. 3, p. 121-122. (7) Gillette, C. P. 1892. Observations upon injurious insects. Colo. Agr. Exp. Sta. Bui. 19. (8) GORHAM, H. S. ' 1897. Biologia Centrali-Americana, Coleoptera. v. 7. , (9) Griffin, H. H. 1897. Results of experiments at the San Juan substation. N. Mex. Agric. Exp. Sta. Bui. 21. (10) Gillette, C. P. 1898. Colorado's worst insect pests and their remedies. Colo. Agr. Exp. Sta. Bui. 47. (11) Casey, T. L. 1899. A revision of the American Coccinellida?. In Jour. N. Y. Ent. Soc, V. 7, p. 71-169. (12) Chittenden, F. H. 1899. Insects injurious to beans and peas. In Yearbook U. S. Dept. Agr., 1898, p. 233-260. (13) Cockerell, T. D. a. 1900. Observations on insects. N. Mex. Agr. Exp. Sta. Bui. 35. (14) Caudell, a. N. 1902. Notes on Colorado insects. In Some miscellaneous results of the work of the Division of Entomologj-. — VI. \J. S. Dept. Agr. Bur. Ent. Bui. 38. (15) Knals, W. 1906. Coleoptera of the Sacramento Mountains of New Mexico. — III. In Ent. News. v. 17, p. 329-333. (16) Fall. H. C, and Cockerell, T. D. A. 1907. The Coleoptera of New Mexico. In Trans. Amer. Ent. Soc, v. 33, p. 145-272. (17) Chittenden. F. H. 1907. Insects injurious to vegetables. New York. (18) Sanderson, E. D. 1912. Insect pests of farm, garden, and orchard. New York. (19) Morrill, A. W. 1913. Entomological pioneering in Arizona. In Jour. Econ. Ent., v. 6, p. 185-195. THE BEAN LADYBIRD. 21 (20) EssiG, E. O. 1915. Injurious and beneficial insects of California. (21) Merrill, D. E. 1917. The bean beetle. N. Mex. Agr. Exp. Sta. Bui. 106. (22) Orton, W. a., and Chittenden, F. H. 1917. Control of diseases and insect enemies of the home vegetable garden. U. S. Dept. Agr. Farmers' Bui. 856. (23) Chittenden, F. H. 1919. The bean ladybird and its control. U. S. Dept. Agr. Farmers' Bui. 1074. 7 p., 3 fig. THE BEAN LADYBIRD IN COLORADO IN 1919. By A. E. Mallory, Scientific Assi'^taiit. LIFE-HISTORY RECORDS. Hibernating adults of the bean ladybird which had passed the winter successfully began to appear in Colorado in 1919 about the middle of June, the first individuals having been found June 16 feeding on beans. A week to 10 days later they began to deposit eggs on the underside of the leaves in clusters of about 40 or more. About 2 weeks later, July 9 and 10, the eggs hatched, and the tiny yellow larvae commenced to feed in a colony near the egg cluster. As they grew older they became separated and did not necessarily con- fine their feeding to the underside of the leaf. As the season ad- vanced, they fed on every part of the plant, blossoms and pods in- cluded. The larvae were present in all stages from the first ap- pearance until at least September 10. On Jul}^ 22, which was 10 or 12 days after the eggs had hatched, pupae were found. During the early part of the season, when foliage is plentiful, pupation takes place on the underside of the leaf. In case the foliage has been extensively destroyed pupation may take place on either side of the leaf or on both sides, 25 to 30 pupae on a single leaf being not un- common. As many as 100 on a single leaf were reported in a case where infestation was heavy. This congregating at the time of pupa- tion seems to be characteristic of the species. Pupae were observed on other plants near beans when the foliage of the beans was almost destroyed. On July 28 adults of the first brood w^ere observed, and by July 30 they were numerous, gradually increasing in number until about September 1 when they seemed to reach their maximum numbers. The first-brood adults are much lighter in color than the hibernating individuals, being a bright yellow at the time of emergence. They gradually become darker, and at hibernating time some are dark red- dish brown, almost if not entirely as dark as the hibernating indi- viduals. Eggs deposited by this new brood were observed August 28 and September 2. It is possible that eggs were deposited by this brood earlier than these dates. 22 BULLETI:N' 843, U. S. DEPARTMENT OF AGRICULTURE. [ There is no definite place in the life history of this species to separate the different broods. Apparentl}^ there are two broods or at | least a partial small second brood for this locality. The fact that | larvae were observed in all stages of development during the entire j season would suggest two broods. Further, the fact that first-brood adults began issuing rather late, and that so few egg clusters were found in late August and September, would suggest only a partial second brood. SrMMARY FOR 1019. Hibernating adults appeared June 16. Beetles present in large numbers by June 25. Beetles began depositing eggs .Tune 25. Eggs began hatching .July 10. Larvae began to transform to pupoe July 22. Adults of first generation issued July 28. (August 28. Egg clusters, second brood, found ISeptember 2 Maximum numbers first-brood adults observed September 1. Adults began to become sluggish September 10. Maximum damage accomplished August 1 to September 1. Second-brood eggs were probably deposited, but not observed, before August 28. INJURY. The bean ladybird so far as observed confines its feeding to beans, and when taken on other plants is never found feeding. The variety of beans seems to make no particular difference as to sus- ceptibility to attack. The adults do not attack the very young 23lants to any considerable extent. This fact is probably due to the small number of hibernating individuals. Furthermore, the over- wintering adults do not concentrate their attack, but move from plant to plant. Thus the damage is less noticeable. Although the adults usually eat entirely through the leaf, they often merely scrape the surface, leaving a network of veins plainly visible. Later in the season as the foliage begins to die they attack the pods, sometimes completely riddling them, but usually eating out small spherical holes here and there along the pod. In the case of canning beans this injury may cause considerable loss, while on seed beans it is not so serious. The percentage of injury by a number of adults is small as com- pared to that caused by an equal number of larvae. The larvae begin their attack on the leaves, invaribly feeding at first on the underside. Instead of eating through they scrape the surface, leaving the skele- ton of the leaf in plain view, although with continued feeding in a limited space they may riddle the foliage. Maximum damage occurred in this locality during the month of August. This is the period when the new adults are feeding along THE BEAN" LADYBIRD. 23 Avith the larvae, and the two broods of hirvse overlap. Damage was generally worst near fences, along ditch banks, and on beans receiving an extra amount of water by accident or seepage. There is a large area in northeastern Colorado devoted to farming and stock raising. This area is known as the Greeley District. Beans of all varieties are grown quite extensively. Several thousand acres are planted to beans every year. A conservative estimate of the damage done to the whole bean crop in that district during the season of 1919 by the bean ladybird is about 5 per cent. This esti- mate at first may seem rather small, but the majority of fields were infested lightly or not at all. In the second place a few fields were more heavily infested than the adjacent or neighboring fields. In those fields coming under our observation the damage by Epilachna corrupta varied from an estimated 25 per cent to 65 per cent. One field in particular, of about 15 acres, was damaged at least 65 per ■cent if not 75 or 80 per cent. No remedial measures were applied in this case. CONTROL. In some of the small truck patches the attack of the bean ladybird was controlled by hand picking. In view of the distribution of the •damage, as stated, clean farming or destruction of winter quarters is suggested as an important measure of control. As to remedial measures, all possibilities were not w^orked out. On July 10 when the first larva? were observed, the following sprays were tested on large plots in a 9-acre field. Right-angle mist-produc- ing nozzles were used. Approximately 80 per cent of the leaves were covered on one side or the other, the remainder varying from a small .amount to none at all. The foliage was heavy. Experiment No. 1. — Lead arsenate, powder, was used at the rate •of 2 pounds to 50 gallons of water, -with 2 pounds of hydrated lime added. Very little if any damage was noted from the spray. A few dead larva^ were found. Experiment No. 2. — Lead arsenate, paste, was used at the rate of 24 pounds to 50 gallons of water. No damage to the plants was noted. Experiment No. 3. — Zinc arsenite, paste, was applied at the rate of 2 pounds to 50 gallons of water. This strength caused no dam- age to the plants. Experiment No. 4- — Bordeaux mixture, alone, formula 3-6-50, was applied to a plot. No damage was done to the plants. This application seemed to be as effective as any of the foregoing during the earlier part of the season, but this plot showed a greater maxi- mum damage after August 1 than did any of the others. The re- mainder of the field and a second field of 21 acres were sprayed with 24 BULLETIN 843, U. S. DEPARTMENT OF AGRICULTURE. the same material and formula as number 1, viz, 2 pounds of lead arsenate, 2 pounds of hydrated lime, and 50 gallons of water. Observations following these applications revealed a few dead larvae. To locate dead larvae was a difficult matter. Our opinion is that many were killed at the first feeding. If not killed then or very soon after, they fed to maturity. Up to August 1 all remedies tried seemed to be equally effective, the amount of damage to the plant being about the same for each plot. Undoubtedl}^ these in- secticides held the beetles and larvae in check. The 9-acre field had an unusually large number of hibernating beetles. After August 1 the damage seemed to increase quite rapidly, and was worse on the Bordeaux plot than on any of the others. A small unsprayed plot was entirely destroyed, most of the injury here being done before August 1. A second spraying two weeks later, about July 25, might have almost if not entirely controlled the attack. In A. B. Owen's field of 9 acres, where spraying experiments were conducted July 10 and 11, the estimated damage was 12.5 per cent. It is believed that approximately 65 per cent of the larvae hatching before July 25 must have been killed by the lead arsenate or zinc arsenite. The number of adult beetles found in this field in early July undoubtedly was large enough to have damaged the crop to the same extent as in the special field mentioned above. This experiment, while not of the type anticipated, demonstrates that the attack of this species can be controlled by the use of either lead arsenate or zinc arsenite. A second spraying between July 25 and August 1 is recommended as very promising in controlling the insect, since it is at about this time that the second brood begins to appear. The two broods overlap, and the damage they do begins to increase very rapidly from this time on. The combination spray consisting of Bordeaux mixture and an arsenical is also a promising experiment. Undoubtedly the Bordeaux mixture, in case it proves a repellent against this insect, will serve as an important fungicide. It will add very little to the expense of spraying, and will possibly increase the yield several bushels per acre by controlling minor fungous diseases. J 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 BULLETIN No. 847 Contrlbatton from the Bureau of Entomolegy L. O. HOWARD, Chief Washington, D. C. PROFESSIONAL PAPER August 9, 1920 ROUNDHEADED APPLE-TREE BORER: ITS LIFE HISTORY AND CONTROL By FRED E. BROOKS, Entomologist, Deciduous Fruit Insect Investigations CONTENTS Page Introduction 1 Historf 2 Distribution 2 Food Plants 3 Distribution as Affected by Native Host Plants 4 Character of Injury S Methods Used in Securing Breeding Material and Rearing the Insects . . 6 Cages Used for Rearing and ObserTing Borers * 6 Page The Egg and Oviposition 7 The Larva 11 The Pupa * . . 14 The Adult 15 Seasonal Phenomena of the Host Trees as an Index to the Time of Develop- mental Changes of the Insect .... 27 Natural Enemies 29 Methods of Control SO Summary 39 Literature Cited 41 WASHINGTON GOVERNMENT PBINTINa OFFICB 1920 UNITED STATES DEPARTMENT OF AGRICULTURE w-^mmm BULLETIN No. 847 •5%llNij>^il#S Contribution from the Bureau of Entomology j^^^IP^ L. O. HOWARD, Chief 0%^^^'^i. Washington, D. C. PROFESSIONAL PAPER August 9, 1920 ROUNDHEADED APPLE-TREE BORER ^r ITS LIFE HISTORY AND CONTROL. By Feed E. Brooks, Entomoloyist, Dechlnous Fruit Insect Investigations. CONTENTS. Page. Introduction 1 History 2 Distribution 2 Food plants 3 Distribution as affected by native liost plants 4 Character of injury 5 Methods used in securing breeding material and rearing the insects- 6 Cages used for rearing and observing borers 6 Page. The egg and oviposition 7 The larva 11 The pupa 14 The adult , 15 Seasonal phenomena of the host trees as an index to the time of develop- mental changes of the insect 27 Natural enemies 29 Methods of control 30 Summary 39 Literature cited 41 INTRODUCTION. In the spring of 1911 a field station of tlie Bureau of Entomology was established at French Creek, W. Va., and a study begun of the roundheaded apple-tree borer in connection with a general project on boring insects attacking deciduous fruit trees. The work was under the direction of Dr. A. L. Quaintance and was placed in charge of the writer, with whom was associated, in 1911 and 1912, E. B. Blakeslee, of the Bureau of Entomology. During the sum- mers of 1915, 1916, and 1917 C. E. Cutright was employed tempo- rarily to assist with the investigation. The field station is located in a hilly, partly wooded region where small orchards, wild seedling apple trees, and native host trees of the borer abound and where the insect itself is plentiful. For rearing purposes and the testing of control measures, 1,000 3-year-old apple trees, of the varieties known as King, Grimes, and York Imperial, were planted in the adjacent locality of Elkins, W. Va., on land leased for the purpose. In addition to the work at the two points mentioned, rearing and life-history studies were conducted at Pick- ens, Weston, and Great Cacapon, W. Va., and at Demorest, Ga., ' Saperda Candida Fabr. ; order Coleoptera, family Cerambycidae. 154187°— 20— Bull. 847 1 BULLETIIT 84*7, TJ. S. DEPAKTMEXT OF AGTtlClTLTUEE. Biltmore, N. C, Wintlirop, Me., and Mimising, Mich. First-hand ob- servations on the species Avere made also in many other localities and more or less original data obtained therefrom. The studies were continued over the period 1911 to 1918. HISTORY. The roundheaded apple-tree borer was first described by J. C. Fabricius in 1787 (1)^ as Saperda Candida-. In 1824 it was rede- scribed by Thomas Say (2) as Saperda Vimttata.^ and by this name it was coimnonly referred to by Harris, Fitch, Walsh, and other early American writers on economic entomology. From 1875 to 1885 Eiley, Lintner, and others recognized the priority of Fabricius's name, and since that time the species has been rightfully designated Saperda Candida. The insect is native to North America and originally fed upon and bred within a limited number of forest trees and shrubs belong- ing to the family Rosace a e. When cultivated orchards of apple, pear, and quince began to be established in the eastern part of the United States the borer soon found its way from the forests into the orchards and did much dam- age to valuable fruit trees. There are many records of serious injury in New York and throughout the New England States, beginning as early as 1825. Apple trees seem to have suffered most ; in some cases entire orchards were destroyed, and the loss of 50 per cent of the trees was not imusual. Felt and Joutel (6) cite numerous historical references showing the widespread and de- structive nature of the insect in the days of pioneer orcharding in this country. In more recent times, as the orcharding interests of the country have developed, losses from this insect have increased rather than diminished. At the present time it is an orchard pest of primary importance throughout a great portion of the apple-growing region east of the Rockv Mountains. DISTRIBUTION. The Imown range of the roundheaded apple-tree borer may be bounded by a line extending from near the mouth of the St. Law- ' Numbers in parenthesis refer to " Literature cited," p. 41. Fig. 1. — Distribution of the roundheaded apple-tree borer (Saperda Candida). EOUISTDHEADED APPLE-TREE BORER. 3 renc© River westward througli Quebec and Ontario to Minnesota, thence in a southwesterly direction through Nebraska, Kansas, and New Mexico to Texas, and thence eastward through Texas, Louisiana, Mississippi, Ahibama, and Georgia to the AtLantic coast. (See fig. 1.) There seem to be no data showing that the general range of the species has been greatly extended by the development of the orchard industry of the country. Within the bounds of its range there are many limited districts where the borer does not occur, or, at least, where it is very uncom- mon. Just why this is true can not be fully explained, but the ab- sence of native host trees and the abundance of those species of wood- pecker which prey upon the borers are two factors which often have much to do with the local scarcity of the pest. Areas of comparative freedom and corresponding areas that are. heavily infested often exist near together for years at a time with little relative change. This occurs in native woods as well as in orchards. The peculiarity may be partly explained by the tendency of the species to colonize or form family breeding centers, far from which the adult females do not habitually wander. FOOD PLANTS. Probably no other tree is so subject to attack by this borer as the quince. Wherever the borer is common it is difficult to succeed with this fruit. Quince trees are usually small and one or two borere can injure greatly or kill a tree in a single year. The habit which the quince has of sending up suckers or sprouts around the central stem tends to give the borers a good chance to work. In the bases of such clumps borers are hard to reach in the worming process and there they may find positions where woodpeckers can not get at them. Apple is undoubtedly the second choice, and probably mountain ash {Sorbus americana) is next in favor. Of the cultivated fruits, quince, apple, and pear are preferred in the order named. Mountain ash, service {AmelancJilei^ canadensis)^ wild crab {Pyrus spp.), haw- thorn {Crataegus spp.), and chokeberry {Aronia spp.) are native hosts which are attacked about in the order stated. There are records in the Bureau of Entomology of the development of this borer in peach, but it is certain that this tree is very rarely attacked. In one instance in West Virginia all the host trees of this borer which grew on a certain tract of woodland and grown-up field were cut and examined to determine the relative extent of infestation of each species of tree. The trees examined numbered 1,483 and the results of the count are given in Table I. It is probable that, in any adjacent locality a considerable variation from these figures might have been found, yet the results of the count showed what is ap- 4 bulletitnt 847, IT. s, depaetme;n-t of ageiculture. parently a constant preference for apple over the other species of host trees present in this instance. Table I. — Relotive numbers of roundheaded apple-tree borers found in different species of host trees at French Creeks W. Va. Trees examined. Number Number. Variety. borers. 194 Seedling apple 85 11 Pear 0 S23 Wild crab 9 405 Hawthorn (Crataegus) 1 50 Service 0 Total 1,483 95 It will be noted from Table I that in this case the 50 service trees (PL I) examined contained no borers. This is far from the rule as regards the service, for in woods where that tree and mountain ash abound tracts are often found Avhere practically every tree is in- fested. (PI. I, B.) Other areas near by of equal size are quite likely to occur where no borers can be found, although the host trees may be just as abundant as where infestation is general. This arises from the fact that under natural conditions families or communities become established and reproduce through many generations within restricted areas. It is probable that adult males fly readily from one breeding center to another, preventing thereby an excess of inter- breeding, but the females do not normally tend to go far from the tree in which they developed, provided other host trees are near. This tendency for infestation to be confined to limited groups of trees is often noted in cultivated orchards. Isolated trees are some- times attacked, however, and there is no question that the female is capable of flying to a considerable distance when impelled by a scarcity of trees in which to oviposit. DISTRIBUTION AS AFFECTED BY NATIVE HOST PLANTS. As has been pointed out, the principal native host trees of the roundheaded apple-tree borer are the service, mountain ash, wild crab apple, and hawthorn. Of these wild hosts service and mountain ash seem to be preferred to the others. It is an interesting fact that the service tree (PI. I, A) occurs over practically the same region of North America as does the insect in question. These wild food plants undoubtedly play an important part in the local and general distribution of the borer. Infestations commonly attributed by orchardists to certain soil conditions, to newly cleared land, or to the hilly contour of the land are in reality usually due to Bui. 847, U. S. Dept. of Agriculture. Plate I. ROUNDHEADED APPLE-TREE BORER. A, Blooming service trees, A melanchier canadensis. At this season of the year borers are undergoing transformation to beetles. B, Exit holes of beetles in clump of young service trees. Bui. 847, U. S. Dept. of Agriculture. Plate 1 1. ROUNDHEADED APPLE-TrEE BORER. Wire-screen cages used on infested apple trees to liold escaping beetles. ROUNDHEADED APPLE-TREE BORER. 5 the proximity of the insect's breeding places, the breeding places quite often consisting of these wild host trees that thrive on account of favorable soil, elevation, or other local conditions. Orchards es- taljlished on newly cleared lands and in hilly locations are more likely to have woods or neglected thickets of wild crab apple, seed- ling apple, or hawthorn growing near to them than are orchards in the more valuable and highly cultivated valley or level lands. The wild host trees that grow in the woods and thickets (PI. I) are usu- ally breeding places for the borer, and the adult insects that develop within them fly to the orchards near by and deposit their eggs. In some localities of the Shenandoah apple region the idea is prevalent that borers discriminate between soils and prefer the shale lands of the hills rather than the clay and loam of the valleys. The greater abundance of borers in hill orchards, however, can be explained by the prevalance in such localities of the wild trees in which they breed (PL VII, C), the soil having only the indirect bearing on the situation that the shale lands favor the growth of service and other wild host trees. The native service tree (PL I) is perhaps the most effective dis- tributor of this insect. In about 25 localities within the States of Maine, Michigan, Pennsylvania, West Virginia, Virginia, North Carolina, Georgia, Alabama, Mississippi, and Florida, where careful investigations were made, the absence, scarcity, or abundance of service trees was accompanied by a corresponding absence, scarcity, or abundance of the roundheaded apple-tree borer. CHARACTER OF INJURY. About 95 per cent of the eggs (PL IV, B, C, E) of this borer are deposited within the bark at the base of the tree trunk. (PL III.) Usually the eggs are within 6 inches of the ground, but occasionally they are placed in a crotch of the tree or even in a branch 10 or 15 feet above ground. The larva (PL V, A, B ; VIII, A), which hatches in early summer, feeds at first on the inner bark, eating out a roughly circular space about the oviposition scar and ejecting string^% saw- dustlike castings of a reddish color through small openings in the outer bark. (PL V, C.) As the larva develops it extends its gal- lery either up or down the tree or transversely with the grain of the bark and before the end of the first season may burrow into the wood. (PL V, A, B.) More frequently, however, it spends the first winter in the inner bark and enters the wood the second summer. The bur- rows, both in the bark and wood, are broad and irregular in form, and, with the exception of a space about the borer (PL VI, D), are packed with digested wood particles (PL VTI, A, C). The borer feeds from about the blooming time of apple in the spring until late 6 BULLETIN 847, U, S. DEPAETIMEi^T OF AGRICULTURE. iA the autumn and continues to throw out castings until it begins the construction of its pupal quarters. (PL VII, A.) Trees of all ages are attacked, but the most severe injury is done to young trees, in which the wound made is gi-eater in proportion to the size of the trunk. (PI. IV, D.) Frequently a number of borers will attack a single tree and girdle it completely or so riddle and weaken the heartwood that the tree will break over at the surface of the ground. It is not unusual to find a dozen borers in one tree, and on one occasion the writer found 25 within an 8-year-old apple tree. Felt and Joutel (6) cite an instance where 30 borers were removed from one tree. Trees severely injured by borers present a sickly appearance, the foliage being sparse and of a pale green color. (PL VI, A.) When of bearing age they are inclined to bloom freely and set heavy crops of fruit, the fruit developing poorly and the trees often dying in an effort to bring the crop to maturity. If any part of an orchard is bounded by woods the first and most severe injury usually occurs among trees near to such woods. METHODS USED IN SECURING BREEDING MATERIAL AND REAR- ING THE INSECTS. In the rearing work connected with this investigation many indi- vidual insects were carried through from eggs to adults in young apple trees planted for the purpose. Larger apple trees were used in which to plant newly hatched borers for rearing purposes. The trunks of some trees were made to support and bring to maturity as many as 25 borers. Each spring a large number of pupa? were se- cured by scouring the roadsides, grown-up fields, and neglected orchards of various localities for small, worthless seedling-apple trees in which the insects were maturing. Such trees were cut near the ground and short sections of the base of the trunk containing the pupae sawn off and taken to the insectary, where they were kept in rearing cages. Many pupae were also chiseled out of trees and placed in small glass vials excluded from the light. About 75 per cent of the pupae kept in the vials developed into normal adults. CAGES USED FOR REARING AND OBSERVING BORERS. In carrying on the work herein described three types of fine- meshed, wire-screen cages were used. The first were small cylinders fitted around the bases of trees in which borers were developing (PL II). These cages were about 15 inches in length, the lower end when in place being sunk in the earth for half an inch and the space at the top between the wire and tree packed with cotton batting. Such cages excluded woodpeckers, imprisoned emerging beetles, and ROUNDHEADED APPLE-TRliE BORER. 7 were useful in other ways in preventing the disturbance of the in- sects. A larger form of cage was made by stretching wire over a light wooden frame 2 by 2 by 4 feet in dimensions. Cages of this type were used to set over apple trees 3, 4, or 5 years of age that had been headed low and pruned in for the purpose. When in place these cages were mounded slightly with earth at the bottom to pre- vent the escape of the beetles and were secured from wind by being attached with screws to posts driven into the ground at the corners. In these cages many beetles were confined over growing trees, and, so far as could be observed, lived lives comparable in length with those in the field. Several other cages of larger size were built over clumps or short rows of young apple trees. Some of these cages were 20 feet in length by 8 feet wide and 8 feet high. They were pro- vided with tight-fitting doors large enough to admit a man and were used for observations on the various stages of the borer and for test- ing control measures. THE EGG AND OVIPOSITION. THE EGG. The egg (PL TV, B, C, E) when first deposited is yellowish white, assuming a darker shade within a few days. It is 3.5 to 4 mm. in length by 1 to 1.5 mm. in width, slightly flattened, both ends tapering to rounded points, the shell tough and plastic, bending somewhat in conformity to the space which it occupies. There is considerable variation in size and shape. THE OVIPOSITION PROCESS. Egg laying usually begins a week or ten days after the female beetles leave the wood. In preparing to oviposit, the female as- sumes an oblique position on the bark (PI. Ill, A) and with her jaws makes a slightly curved slit in the bark 4 or 5 mm. in length, and usually extending parallel with the grain of the bark. (PI. IV, A.) After the incision is completed, the beetle turns, inserts the tip of the ovipositor into the opening, and with considerable effort forces it into the tissue, usually between the bark and wood. (PL IV, B, C.) The ovij)ositor is inserted at about the center of the slit made with the mandibles and is extended under the bark in a direc- tion at right angles to the slit. The egg is placed with the end toward the slit and from 1 to 2 mm. from it at the nearest point. After the ovipositor is withdrawn a small mass of clear, gelatinous liquid is ejected into the hole, which dries and seals the egg chamber. Two or three minutes are spent in making the initial slit and twice that time in inserting the ovipositor, laying the egg, and sealing the 8 BULLETIN 847, V. S. DEPARTMENT OF AGRICULTURE. opening. Several eggs are usually deposited at a time within one tree (PL IV) , quite often the second slit in the bark being made joining and in line with the first. When the bark over eggs is peeled off the eggs adhere to the bark rather than to the wood. (PL IV, B, C.) After the batch of eggs is deposited the beetle may crawl up the trunk to the branches or may move away on the ground for a little distance and then take wing. TIME OF DAY DURING WHICH EGGS ARE LAID. Most of the eggs are deposited during the hotter part of warm, sunny days. No evidence was obtained that oviposition ever takes place at night. In one case a female was fomid ovipositing at 5.30 a. m. and others were several times observed laying eggs near twilight in the evening, a time of special activity with both sexes of beetles. PLACE OF OVIPOSITION. As has been stated on another page, probably 95 per cent of the eggs of the roundheaded apple-tree borer are deposited in the trunks of trees within 6 inches of the ground ; usually they are not more than 1 or 2 inches above the soil and quite often they are on a level with (PL III, C) or slightly below the surface. (PL III, A, B.) In rare instances eggs are deposited in the crotches of trees, around the edges of cavities in the trunk or larger branches, and even in small branches high up in the trees. In one case the writer found two larvae working in a branch 15 feet above the ground. Usually woodpeckers remove the borers that begin operations aloft in the trees. Possibly the beetle's habit of ovipositing close to the ground has evolved from the greater dangers attendant upon the higher locations. Wlien the female beetle is ready to oviposit she usually crawls from among the branches downward along the trunk. In descend- ing, if an obstacle of any kind is encountered she may pause and oviposit in the bark above it. This habit accounts for borers occa- sionally found in the crotches of trees. The writer found numerous eggs and larvae above burlap bands which had been placed around the trunks of apple trees for trapping codling moth larvae. When similar bands were placed on trees in cages the beetles laid more eggs above the bands than at the ground. In one test of this kind, com- prising 5 trees and 5 female beetles, 36 eggs were laid above the bands and only 15 at the ground below the bands. These bands were attached around the trunks about 15 inches up from the ground. The statement has been made that the females prefer to oviposit in the trunks of trees that are surrounded and shaded by weeds. Bui. 847, U. S. Dept. of Agriculture. Plate III, ROUNDHEADED APPLE-TREE BORER. A, Female beetle splitting the bark of a young apple tree just below the surface of the ground pre- paratory to depositing an egg. Natural size. B, Female beetle placing an egg in the tree just below the surface of the ground. Natural size. C, Female beetle ovipositing at the surface of the ground. Slightly enlarged. Bui. 847, U. S. Dept. of Agriculture. Plate !V. ROUNDHEADED APPLE-TREE BORER. A, Oviposition scars in bark of young apple tree. Natural size. B, Inner surface of bark peeled from young apple tree, showing position of eggs. Natural size. C, Eggs in natural position. Much enlarged. D, young borers attackiug lower trunk of young apple tree. E, Eggs. Much enlarged. ROUNDHEADED APPLE-TREE BORER. 9 litter, or water sprouts, and that trees so surrounded are more likely to be injured by borers than those having trunks exposed to the full light. Observations made by the writer, however, indicate that the beetles prefer to oviposit in the sunlight rather than in the shade. In one case 609 borers were removed from an orchard of 6-year-old apple trees which was located on a hillside having a general slope to the southwest. The trunks of the trees were long and one side re- ceived the unobstructed sunlight during a greater part of the day. The oviposition scars in which the borers hatched were found and those situated on the sunny side and shady side of the trunks were counted separately. As nearly as possible, each side was made to include half the circumference of the trunk. Of the 609 ovijoosition scars which were located, 393 were on the sunny or exposed side and 216 on the shaded side. The fact that about 65 per cent of the eggs had been laid on the more exposed side of the trees indicates that mulches of straw or hay placed around trees or the shading of the trunks by low branches does not attract the beetles, but rather repels them. OVIPOSITION IN CAGES. Other workers with this insect seem to have had but little trouble in inducing caged fem\ale beetles to oviposit in apple twigs, in sec- tions of apple branches set in the ground, and even in apple fi'uits. The writer, however, has never been able to get a normal frequency of oviposition in any but growing trees. At one time 15 pairs of beetles were confined separately in roomy cages for a period of 17 days. Fresh apple branches, about an inch in diameter, were set in the ground daily in each cage. Although this was kept up from May 31 to June 17, a time when egg laying in the field was at its height, onl}^ 17 eggs were secured in the branches and 10 of the 15 females failed altogether to oviposit. The cages with the beetles were then removed and placed over small growing apple trees, whereupon oviposition began freely and at once. Attempts to induce caged females to oviposit in other than growing trees were made frequently, but never with entire success. OVIPOSITION PERIOD. As an example of the period of time over which eggs are de- posited in a given locality, observations made in 1914 may be cited. During that year the first female issued on May 23. Eight days later she paired with a male and on June 4 laid her first eggs. No record was obtained of the last egg of this individual but other females continued to oviposit until August 1, the entire egg-laying 154187°— 20— Bull. 847 2 10 BULLETIN 847, U. S. DEPARTMENT OF AGRICULTUEE, period for the year covering 58 days. It is not unusual to find un- hatclied eggs in the field in the latitude of West Virginia from Au- gust 10 to 15, although none have been found after August 15. From these and from considerable other data obtained, it ai^pears that nor- mally egg laying continues in a given locality from 50 to 60 days, although perhaps no single individual oviposits over so long a period of time. INDIVIDUAL EGG CAPACITY. The number of eggs normally deposited by a single female beetle is not great, the average number having been found by different in- vestigators to vary considerably. Felt and Joutel (6) mention 10 as being about the quota for one individual, while Becker (14), mak- ing observations in the Ozark Mountain regions, found the number to range from 16 to 93, with an average of 40.8 per female for five individuals. The writer obtained egg counts from 15 individuals, the figures beinjr shown in Table II. Table 11.— Individunl eyg capacity of feuKiIc beetles of the roundJieaded apple- tree borer at French Creek, W. Va. Year. Xumber of females. Number of eggs. Average number of eggs per female. 1911 3 52 17 20 22 13 20 193 17.3 1914 17 1914 20 1914 22 1912 . . 13 1912 20 1914 ; 27.6 Total . .... 15 337 Average number of eggs per beetle 22. 5 Table II shows that the minimum and maximum numbers of eggs obtained from 15 beetles were 13 and 27.6, the average being 22.5. Since these records were obtained in several different years under conditions that were approximately normal they probably represent fairly accurately the average number of eggs laid by females in the locality mentioned. PERIOD OF INCUBATION OF EGGS. The time required for the eggs to hatch has been variously stated at from 8 days to 3 weeks. The writer noted the period in 8 in- stances, the data for which are set forth in Table III. ROUNDHEADED APPLE-TREE BORER. 11 Table III. — Period of incuhation of the roundheaded apple-tree borer at French Creek, W. Va. Pates on which eggs were laid. Dates on which eggs hatched. Number of eggs. Number of days re- quired to hatch. June 14 June 28 14 June 12 do 16 June 9 June 25 16 Do June 26 17 June 10 June23 13 June 13 June 29 16 Do June 30 17 Do July 2 19 Average period of incubation 16 days. As is shown in Table III, the minimum and maximum periods of incubation were 13 and 19 days, the average for the 8 eggs being 16 days. Evidence was obtained that hatching is retarded by low and accelerated by high temperatures. THE LARVA. The larva (PI. V, A, B) is a cream-colored, footless grub, with brown head, blackish mandibles and a conspicuous patch of small, brown tubercles on the posterior half of the broad, flattened dorsal surface of the first thoracic segment. The intersegmental con- strictions are deep and the dorsal and ventral surfaces of the first seven abdominal segments are elevated and roughened. The sides of the body are sparsely covered with short, stiff hairs. When full grown, the length is from 30 to 40 mm., the females being consider- ably larger than the males. According to Becker (14) there may be as many as six larval instars. DEVELOPMENT AND FEEDING HABITS OF THE LARVA. The behavior of the larva varies as affected by individual char- acteristics, difference in size, vigor, and species of host trees, and dif- ference in localities, so that no one description of the larval develop- ment will apply to all. In the latitude of West Virginia the activi- ties and growth of the larva when feeding under normal conditions in apple are about as follows : The larva begins to feed at once after leaving the egg and soon eats out a broad, irregular, usually more or less circular burrow around the point where the egg was laid. At this stage of its life growth is rapid and the borer soon forms a broad, elongate gallery under the bark which may extend in any direction away from the point first attacked. As winter approaches there is some tendency to burrow downward to or beneath the soil, but this is by no means general. At first the feeding is all in the bark and the point of injury usu- ally shows from the outside as a dark, slightly depressed spot from 12 BULLETIN 847, U. S. DEPARTMEN^T OF AGRICULTURE. which castintrs are ejected arid from which a small quantity of sap often flows. In some cases the borer burrows into the wood the first season, but usually it does not enter the wood until the succeedin-i' spring. In small trees the galleries penetrate to the heart, but iu old trees they are seldom extended more than an inch beneath the inner bark. The burrows in the wood, like those in the bark, are broad and irregular in shape and usually extend both above and below the surface of the ground. In the northern part of the in- sect's range a greater proportion of the feeding seems to take place beneath the ground. The writer found larvse in Maine burrowing downward in the roots to a distance of a foot or more from the base of the trunk, a depth which does not seem to be reached in the South. Many 1-year-old larvae were also found in Maine that had not yet penetrated into the wood but were still feeding in the bark near the old oviposition scars. In all their feeding larvse keep an open space about themselves, to allow of free movement, but pack the balance of their burrows with wood fragments. Strings of red- dish-brown castings are also thrown out from the tree through small openings in the bark. (PI. V, C.) In the late summer and autumn preceding the spring during which pupation is to take place, the larvfe excavate galleries leading up the trunk of the tree a short distance beneath the bark. (PI. VHI, A, B.) At the upper end of this gallery the pupal chamber is formed by slightly enlarging the circumference of the opening and curving the upper end outward to the inner bark. (PL VII, A.) The curved upper end is packed lightly with wood dust and a con- siderable space in the gallery below the pupal chamber is filled with short, excelsior-like strings of wood torn from the walls of the opening. (PI. VIII, A.) The space for the pupa is often 2 inches or more in length and both the larvse and pupa? when occupying it recede or advance when disturbed, evidently a provision for escaping wood- peckers. The pupal quarters usually are practically completed in the autumn but the larvae add finishing touches in the spring before they pupate. In small trees the exit holes at the upper end of the pupal chambers are usually within from 4 to 8 inches of the ground, but in large trees it is not unusual to find the place of exit at the terminus of a gallery extending upward from the ground to a distance of 18 inches or 2 feet. Just why the pupal quarters should be made higher in large trees than in small trees does not seem to have been determined. Wintering larvse begin activities early in the spring and continue to feed until stopped by the cold weather of winter. Probably the annual feeding period in the South is much longer than in the North, Bui. 847, U. S. Dept. of Agriculture. Plate V. ROUNDHEADED APPLE-TREE BORER. A, Borer, first summer in tree. Natural size. B, Second summer in tree. Natural size. C, Cast- ings being ejected from tree by borer. Bui. 847, U. S. Dept. of Agriculture. Plate VI. ^, ^^* 4|B^ ss?? "• v4;"' Ml ROUNDHEADED ApPLE-TREE BORER. A, Young apple tree dyinf; from injuries caused by roundheaded apple-tree borers. B, Trunk of young apple tree marked by beak of woodpeckers searching for borers; the larger wounds show where borers have been removed. C, Adult roundheaded apple-tree borers; male above and female below. Slightly enlarged. D, Borer in apple tree; showing cleared space maintained in burrow to allow of free movement of body. Natural size. Bui. 847. U. S. Dept. of Agriculture. Plate VII. ROUNDHEADED APPLE-TREE BORER. A, I'upffi in natural position in tree. B, Castings of borers at base of young apple tree. C, Borers in young service tree. D, Exit holes of beetles. Dept. of Agriculture. ROUNDHEADED APPLE-TREE BORER. Natural size. C, Beetle resting by exit hole. Natural size. ROUNDHEADED APPLE-TREE BORER. 13 FORM OF BURROWS IN PEAR DIFFERS FROM THAT IN APPLE. It was frequently noted that the borers in pear trees formed dif- ferently shaped burrows from those made in apple. In pear tlie burrow is much more elongate, often being a slender gallery 6 or 8 inches in length and extending around the trunk, sometimes almost or entirely encircling trees several inches in diameter. As the larva? in pear near maturity they enter the wood and pupate much as in apple. PERIOD SPENT BY LARVA IN THE TREE. Ever since the roundheaded apple-tree borer began to attract the attention of entomologists there has been some disagreement as to the number of years spent by the larva in the tree. Practically all writers have agreed that the life cycle requires either two or three years for completion. Most of the well-known textbooks on general entomology, as well as the systematic treatise on this particular species, give three years as the life period. Comstock (4, p. 573) says "It requires nearly three years for this larva to attain its growth." Smith (5, p. 209-210) , speaking of the larva, says " In the spring of the third year [it] changes to a beetle." Felt and Joutel (6) give a three-year period in the tree. Saunders (8) says " It is generally conceded that the larva is three years in reaching maturity." Sanderson (12) says " The third spring the larvae trans- form to pupae." Slingerland and Crosby (13) state, "It is generally believed that it requires three years for this apple-borer to complete its life cycle." Lutz (15, p. 359) says, " From egg to adult takes three years." Chittenden (T) gives a three-year life cycle. O'Kane (11) says "The larva requires three years for maturity." Both Smith (10, -p. 52-51) and Patch (9) give a three-year larval period. Becker (11) says ^^Saperda Candida has a two-year life cycle in the Ozarks," but points out that " There seems to be some indication that occasionally a larva may require three years for its development." The present investigation has shown that the length of the life cycle averages longer in the North than in the South and also that this period may vary several years in length in a given locality. Table IV shows the years required for 121 insects to reach maturity at French Creek, W. Va. Table IV. — Period of life cycle of the rmindheaded apple-treG borer at French CreeJc, W. Va. Year beetles issued. Niunber years in tree. 1 2 3 4 5 1913 . ...'.. . . 0 0 2 0 0 9 11 7 36 40 0 2 1 10 2 0 0 0 1 0 0 1914 0 1915 0 1916 0 1917 - - - 0 Total 2 1.7 10.3 85.1 15 12.4 1 0.8 0 14 BULLETIlSr 847, U. S. DEPAETMEXT OF AGRICULTURE. Table IV shows that out of 121 individuals, 2 issued from the wood as beetles the next year after the eggs were deposited, 103 is- sued in two years, 15 in three j^ears, and 1 in four years. Records of the exact number of days elapsing from the dej)osition of the eggs to the issuing of the beetles were ascertained in a number of instances. These records are shown in Table V. Table V. — Number of days hettccen deposition of egg and emergence of beetle of roimdheaded apple-tree borer at French Creek, W. Ya. Date egg was deposited. Date beetles issued from wood. Number of beetles. Male. Female. Number of days from egg to beetle. Jime IS, 1913 . Do Do June 15, 1914. June IS, 1914. June 8, 1914 . . June 12, 1914. June 18, 1914. June IS, 1913. June IS, 1914 . June 16, 1914. June IS, 1914. Jimell,1914. June IS, 1913. June 11, 1914. June 16, 1914 . Total... June 1,1915.. June 14, 1915. Junes, 1915.. May 24, 1916. do..:.... May 25, 1916.. do May 26, 1916. May 27, 1916.. do May 2S, 1916.. do do do do May 29, 1916.. 713 726 720 708 705 716 712 707 1,073 708 711 709 716 1,074 716 712 10 10 As is shown in Table V, 1 male spent 1,073 days in undergoing development within the tree and 1 female 1,074 days, a 3-year period for each. All the others reached maturity in 2 years, the 9 males requiring on an average 710 days from Qgg to adult and the 9 females undergoing the same transformation in an average of 714 days. Records obtained at Winthrop, Me., show that of 24 individuals, none matured in 2 j^ears, 6 issued as beetles in 3 years from the egg, and 18 required 4 years to develop from eggs to adults. No definite records were obtained for individuals requiring longer than 4 years for development and yet observations were made which indicate a 5-year period in some cases. Observations made at Biltmore, N. C, indicate a 2-year period with most individuals in that locality. THE PUPA. The pupa ("PI. VII, A), when first formed, is soft and delicate, the color being similar to that of the larva. Within a few days it turns slightly yellowish, the eyes soon take on a dark color, and later the whole body becomes mottled with brown and blackish markings. The pupa occupies a vertical position in the cell and measures from 18 to 25 mm. in length, the females being much longer and more robust than the males. ROUNDHEADED APPLE-TREE BORER. 15 PERIOD OF PUPATION. .At French Creek, W. Va., practically all the individuals pass through the pupal stage during the period from April 15 to May 15, although the time of pupation and the duration of the stage depend very much on weather conditions. The earliest pu^^a was found in the field April 12, 1913, and the latest May 20, 1915, although undoubtedly a few may sometimes occur before and after these dates. In the locality mentioned, the entire pupal period for any one year has not been found to extend over 30 days, the pupal stage for a single individual averaging about 20 days. All transforming individuals were found in the pupa stage at Demorest, Ga,, on May 1, 1915 ; at AVinthrop, Me., on June IT, 1916; and at Munising, Mich., on June 20, 1917. As has been stated, the pupse are sensitive to temperatures, warm weather accelerating and cold weather retarding the changes. After transforming to the a-dult stage the beetles usually remain within the pupal chamber from 5 to 10 days, the length of this period, too, depending on weather conditions. THE ADULT. The adult (PI. VI, C ; VIII, C) is a handsome, elongate beetle, the males averaging 15 mm. in length and the females 20 mm. The back is cinnamon brown with two broad white stripes extending the full length of the body; the front of the head and underparts are silvery white and the legs and antennae gray, changing to brownish at the extremities. The antennae of the males are slightly longer than the body and those of the females slightly shorter than the body. In escaping from the wood the beetles gnaw round exit holes through the bark at the upper end of the pupal chamber, the holes ranging from 5 to 8 mm. in diameter, the larger ones being those from which females issue (PI. VII, D). PERIOD OF ACTIVITY OF BEETLES. The statements of other investigators regarding the length of time that the beetles of the roundheaded apple-tree borer are on the wing indicate that in some places this period may be of longer duration than in any of the localities where the present writer has made observations. Becker (14), in the summary of his paper, says that in the Ozarks pupation begins in the latter part of March and may continue until the middle of June, and on another page speaks of larva? that pupated as late as July 11, the inference being that this was under normal conditions. The beetles are said by the same author to be on the wing in the Ozarks from the third week of April 16 BULLETIN 847, U. S. DEPARTMENT OF AGRICULTURE. until perhaps as late as the last of August. These statements indi- cate that in the Ozarks beetles continue to issue from the wood over a period of approximately 100 days. Chittenden (7) states that oviposition has been observed from June to September in a single locality (Lawrence, Kans.) and says that at Albany, X. Y., beetles have been observed in the trees as early as April. Felt and Joutel (6) cite statements of observers giving the months of May, June, July, and August afe the time when beetles are abroad. All the rearings of the present writer indicate that the beetles issue from the wood over a much shorter period than has been found by the writers referred to above. The longest periods covered by the emergence of the beetles were at French Creek, W. Va., where, in 1916, 214 beetles issued from May 20 to June 18, a period of 30 days, and in 1917, when 118 beetles issued from May 25 to June 23, a period of 30 days. No beetles in either year issued before the first dates or after the last dates named. In all the rearings of which the dates of issue were kept, including 772 beetles, the first to issue was at Demorest, Ga., on May 8, and the last at Munising, Mich., on July 23, the interval between these two extreme dates beino; 78 days. Table VI presents the data which have been accumulated rela- tive to the time of emergence of beetles in several localities. Tabi^e VI. — Periods during icJiich beetles of the roundheaded apple-tree iorer issued from the ivood in, different localities. Locality. Num- ber of bee- tles. First and last emergence. Num- ber of days. Periods over which beetles issued from wood. Year. May. June. July. 5 10 15 20 25 30 4 1 9 14 !! 24 29 4 9 14 19 _ 24 29 1911.. 1912 French Creek, W. Va . ... do 11 27 95 116 5 244 118 21 35 13 16 24 1 28 6 2 8 2 May 23 - June 1... May 14-June 6... May 12-June 6... May 22 -June 8 . . May25-Junel4.. May20-Junel8.. May 25- June 23.. Mayl8-May28.. May 20- June 2... May 20- June 6... May 28- June 5... May2S-June9... June 21 10 24 26 18 21 30 30 11 14 18 9 13 1 13 7 6 14 3 1913 do do ■ 1914.. 1915.. do •• .. MM _^__^ 1916.. do 1917 . .do -I" 1918 do ■ 1 1 1914.. 1914 Weston, W. Va Great CacaTwn, W. Va. Elkins, W. Va Pickens, W.Va Pemberton, N.J Winthrop, Me Demorest, Ga Biltmore, N.C Winthrop, Me Munising, Mich Total number beetles.. ■ -»— i 1914.. 1914.. 1914.. ;; •■ 1 1 1914.. 1915. . 1915.. 1916 June 18-June 30. May8-May 14... Mayl5-May20.. June20-Jaly3... Jaly23-Jaly25.. •• •• 1917.. ■■ 772 LENGTH OF LIFE OF INDIVIDUAL BEETLES. Eight female beetles whose emergence and death were noted in 1911 and 1912 lived, respectively, 27, 31, 37, 41, 41, 44, 46, and 46 EOUNDHEADED APPLE-TREE BORER. 17 days. In 1912 the first beetles issued from the wood on May 14 and eggs were still being deposited on July 1, 48 days after the first beetles appeared. In- 1913 beetles were observed on the wing from May 12 to July 19. a period of 68 days. In 1914 a female issued on May 27 and died July 24, living 58 days. In the same year a male and female were alive on August 6, 76 days after the first beetle issued. In 1917 the first beetle left the wood on May 25 and the last beetle of the year died August 10, beetles thus being on the wing for 77 days. In 1918 two females were observed to be alive and active 61 days after leaving the wood. The beetles referred to above were in all cases kept in roomy wire-screen cages over small, growing apple trees, and it is presumed that their life periods extended over the normal term. FEEDING HABITS OF BEETLES. The beetles feed to a considerable extent upon both tender and tough bark of twigs and branches and upon leaf stems and leaf ribs, and they frequently chew ragged holes through the tissues of the leaf. (PI. IX, A, C.) They were observed often working with their mandibles at the castings ejected from trees by larvae of their own kind and were seen occupied in a similar manner with damp soil; this was probably for the purpose of obtaining water. One female beetle kept alone in a cage over a young apple tree lived for 58 days, and after her death a careful measurement was made of the leaf and bark surface over which she had eaten. The total area eaten was found to be 6.9 square inches. In another instance two male and three female beetles, which had just issued from the pupal quarters, were placed in a roomy cage over a young apple tree that had been sprayed just before with lead arsenate at a strength of 3 pounds of the paste to 50 gallons of water. Two of the beetles died the first day, one died on the second day, one on the third, and the other, a female, died on the ninth day. All apparently succumbed to the poison, as" there was no mortality among beetles caged at the same time over unsprayed trees. Death occurred to all the beetles confined over the sprayed tree before any eggs were deposited. It was noted frequently that beetles differed individually in the amount of feeding done immediately following their emergence from the wood, some proceeding to feed at once and others waiting several days. It is probable that the female referred to above, which lived nine days over the sprayed apple tree, did no feeding until a short time before her death. COPULATION. Copulation may take place soon after the beetles issue from the pupal chambers or it may be deferred a week or 10 days, the time 154187°— 20— Bull. 847 3 18 BULLETIN 847, U. S. DEPARTMEN-T OF AGRICULTURE. of pairing seemingly depending about as much upon the volition of one sex as the other. Ne^iy emerged males occupying cages in com- pany with females have been observed to wait several days before paying any attention to the females. In other cases they have begun courtship on the day following that of their emergence. The females usually repel the males for several days, but will sometimes receive them within an hour after quitting the pupal quarters. Evidently some individuals of both sexes remain in the wood until they are sexually mature, while others issue before the sex instinct has developed. The act of copulation usually lasts several hours and is repeated at frequent intervals so long as both sexes live and are active. Pairing was several times noted after the participants had been on the wing from 30 to 40 days. Females confined by themselves were observed to engage in a performance evidently to attract males. They would occupy the upper surface of an exposed leaf and thrust out the ovipositor to its utmost length and then wave it about while it was being gradually drawn in. A few minutes later the ovipositor would be again ex- tended and then drawn in and so the act would continue for an hour or more. Apparently a scent or influence of some kind was being discharged as a sex attraction, but when females so engaged occupied outdoor cages no wild males of the locality were observed to come to the cages. DAY AND NIGHT ACTIVITIES OF THE BEETLES. Early writers on this insect described the beetles as being active nocturnally and secreting themselves by day. The beetles" were sup- posed to issue from their pupal cells and deposit their eggs exclu- sively by night. The reverse of this habit, however, has been found more nearly true. All the beetles issue from their exit holes by day, usually during the forenoon, although a few continue to come forth during the afternoon hours. No evidence was obtained that ovi- position ever takes place in the darkness, although male beetles are occasionally on the wing at night. There is a period of activity in the evening just before twilight when both sexes are especially in- clined to flight, but as darkness comes on most of the beetles settle among the branches and remain quiet until the light of the morning. Observations made at night with electric flashlights indicated that the normal habit is to rest in one place through the night, but that occasionally the beetles move about in the darkness. DO THE BORERS DIFFERENTIATE BETWEEN VARIETIES OF APPLE? Orchardists often observe what appears to be a preference on the part of the roundheaded apple-tree borer for certain varieties of KOUNDHEADED APPLE-TREE BORER. 19 apple. Individual trees or blocks of one kind of apple will be at- tacked year after year much more extensively than those of other varieties. Becker (14) concludes from experiments that the borer does not discriminate between varieties and that the preference which is often indicated is merely a matter of propinquity. During the present investigations observations were made bearing on this point over a period of five years in the experiment orchard at Elkins, W. Va., As has been stated, this orchard contained only three varieties, namely: 310 King, 341 Grimes, and 341 York Imperial, the block of Grimes occupying a space through the center of the orchard. All the rows of the three varieties abutted impar- tially at one end against an older and heavily infested orchard. There were no conditions within or surrounding the orchard that would appear more favorable for the attack of one variety than another, except that female beetles in entering the orchard to ovi- posit might be expected to alight more frequently on the outer trees. During four of the five years over which counts were made, however, the Grimes in the center were much more severely at- tacked, practically 50 per cent of the 1,639 borers removed from the trees being found in this block. The King trees, although fewer in number, were second in point of attack, and the York Imperial trees suffered least. This ratio of attack, as may be seen from Table VII, was constant for the years 1913, 1914, 1916, and 1917. In 1915, which was the year of lightest infestation, the York Imperials were first in point of attack, the Grimes second, and the Kings third. Takle VII. — Relative extent of infestation bp the roundheaded apple-tree borer of three varieties of apple for a period of five years. Variety of apple. Year. King. Grimes. York Imperial. Number of borers. Per cent. Number of borers. Per cent. Number of borers. Per cent. 1913 49 80 8 223 121 30.3 33.5 8.3 35.4 23.7 77 110 43 283 303 47.5 46.0 44.8 44.8 59.3 36 49 45 125 87 22.2 1914 20 5 1915 46.9 1916 19.8 1917 17 0 Total 481 29.3 816 49.8 342 20.9 It is entirely possible that the results which are shown in Table VII are accidental, and yet it must be confessed that, aside from varietal preference on the part of female beetles while ovipositing, there is no apparent way of accounting for the almost constant maxi- mum attack of Grimes and minimum attack of York Imperial. 20 BULLETIlSr 847, U. S. DEPARTMEISTT OF AGEICULTURE. DISTANCE OF FLIGHT OF FEMALE BEETLES DURING OVIPOSITION. The probability that the female beetle during her egg-laying ac- tivities does not normally wander far in search of host trees has been suggested by the fact that the trees containing the larvse are usually found in somewhat restricted groups. This grouping of the borers occurs not only in orchards but in the woods as well, and indicates that where host trees grow near together the adult females during oviposition are not inclined to fly far from the trees in which they develop. In an effort to obtain data bearing on this point several experiments were carried out in West Virginia, which are described below. THE DAENATX OECHAED. This orchard contained 537 apple trees ranging in age from 4 to about 20 years. In the summer of 191-i it was found to be badly in- fested with roundheaded apple-tree borers, and the trees were gone over carefully after the eggs of the current season had hatched, all the borers being removed and counted. The trees contained 141 borers, 106 of which had only recently hatched. The orchard was surrounded by pasture lands and woods in which grew an abundance of seedling apple, wild crab apple, hawthorn, and service trees. Within a strip 600 feet in width surrounding the orchard these out- lying trees were also examined and all the borers removed and counted, the number of borers found being 95. This operation was rej)eated annually for a period of 4 years, it being ob\dous that if borers were not allowed to breed within the area all the young borers found within the orchard after the first year would necessarily have hatched from eggs deposited by female beetles which had flown into the orchard from outside the 600- foot strip. In the second year of the experiment (1915) an examination of the orchard trees showed that female beetles had crossed the 600- foot strip and deposited 56 eggs. The third year (1916) only 1 egg was deposited in the orchard. The fourth year (1917) beetles flew across the boundary strip and deposited 44 eggs in the orchard trees. While this was a considerable reduction from the number of eggs deposited annually in the orchard before the experiment began, still it showed a rather general tendency on the part of the female beetles to fly at least 600 feet in searching for trees in which to oviposit. In figure 2 are given plats of the Darnall orchard showing the location of infested trees and the number of borers found at each of the four annual examinations. ROUNDHEADED APPLE-TREE BORER. 21 ® (D O © © © ®1® 0 (|) © ® ®© © ©^ @ ®® © © ® ©^ ® © ® ® ®® ®®® ^ (0®®® ® 0® 0 © ©0 ®0 o C3)@@ Fig. 2. — Saperda Candida. Plats of Darnall orchard illustrating distance of flight of female beetles during oviposition. Circles represent location of infested trees and the figures within them show number of borers found. No beetles were allowed to develop within 600 feet of the orchard. Plat 1 : Position of infested trees and number of borers found at first examination (1914). Trees contained 106 young borers and 35 from eggs of previous seasons. Plat 2 : Position of 56 borers devel- oping from eggs deposited in 1915 by beetles that had flown into the orchard from outside the 600-foot strip. Plat 3 : Position of one egg deposited by female which entered the orchard from outside the 600-foot strip in 1916. Plat 4 : Posi- tion of 44 eggs deposited by beetles which entered the orchard from outside the 600-foot strip in 1917. 22 BULLETIN 847. V. S. DEPARTMENT OF AGRICULTURE. THE PAGE OKCHAED. The Page orchard, like the Darnall orchard, was found in 1914 to be heavily infested with roimdheaded apple-tree borers. The trees within the orchard and within a strip 300 feet in width sur- rounding the orchard were cleaned of borers. The orchard contained 464 trees from which were removed 290 borers, 254 of which were from eggs of the current season. This orchard was surrounded en- tirely by pasture lands over which grew scattering seedling apple, crab apple, and hawthorn trees in which many borers were develop- ing. The second annual examination, which was made in August, 1915, showed that 55 eggs had been laid in the orchard trees. One Fig. 3, A. — Saperda Candida. Plat of Page orchard illustrating distance of fliglit of female beetles during oviposition. Circles represent locations of infested trees and the figures within show number of borers found. Orchard surrounded by 300-foot strip cleaned of borers. Plat 1 : Infested trees and number of borers found at first examination (1914). Orchard contained 290 borers, 254 of which developed from eggs of the current year. borer had been overlooked in the orchard during the examination of the previous year and this had developed into an adult female, as was apparent from the size of the exit hole, and near to the tree from which it issued two trees were found containing, respectively, 4 and 9 young borers. In another part of the orchard a group of 7 trees contained 42 young borers. This group of infested trees was near the outer border of the orchard, and 275 feet distant another fresh female exit hole was found in a seedling apple growing in the pasture field, the author of which had been overlooked the previous year. It seemed probable that this beetle had flown to the orchard and that the two females overlooked the previous year were respon- sible for all the eggs which were deposited within the orchard in 1915. ROUNDHEADED APPLE-TREE BORER. ' 23 The third year (1916) 38 eggs were distributed among 10 of the orchard trees, all evidently having been laid by females that flew into the orchard over the 300-foot strip. "*" 175 feet Exit (D ® Fig. 3, B. — Saperda Candida. Plat at Page orcliard illustrating distance of flight of female beetles during oviposition. Circles represent locations of infested trees and the figures within show number of borers found. Orchard surrounded by 300-foot strip cleaned of borers. Plat 2 : Xumber of borers found in second examination (1915). Fifty-five borers present from eggs of current season. Crosses indicate where two female beetles issued that had been overlooked as borers during the previous annual examination. 0) © 0® 1 -* 1 1 Sj / s 1 'S / \ V ^ I > \ 1 / \ j \ \ / \ / 1 / ) 1 / > f > I 1 s / s 1 / / \ / \ 1 N , / f \ / > / / -\ ' f V> k \ > 1 s \ ^ / \ \ \ / ' / \ \ \ ^ y V / f \ 1 V 1 V I / \ \ i J \ / \ \ / I. _ _ lA J _ . l\ Lj J _ Fig. 4. — Saperda Candida. Diagram showing tendency of female beetle to refrain from long flights during oviposition. A, Number and distribution of borers found in or- chard after liberating 124 female beetles during three separate years on row 16. i}j Number and distribution of borers found after libeiuting 9 females on row 29. All the foregoing data on the flight of the female point to a con- stant tendency on her part to deposit eggs near the place of her de- velopment. They also afford good evidence that the female beetle is capable of flights of considerable distance when impelled by any special desire. In one case when a female was liberated in the manner described above, she immediately took wing and arose to a height of 30 or 40 feet and then disappeared in the direction of a tract of woods about EOTJE'DHEADED APPLE-TREE BORER. 27 500 yards distant. That such flights are unusual, however, is indi- cated by all the evidence that could be gotten. FEMALES LESS PKECOCIOUS THAN MALES. The females are not only less active in flight and more sluggish generally than the males but are regularly two or three days behind the males in issuing from their pupal quarters in the wood. This constant tendency on the part of females to be slower than the males in emergence is illustrated by figure 5, 1^ 28 Z7 26 ZS 2* ZS 22 Zl ZO /9 /S fr /6 /5 /4 /S /2 // /O 9 6 7 6 S 4 3 2 / xDAYS or nONTM m It iO It 32 Zi l-t il X 2T Ba IS 30 3/ 1 Z 3 4 S 6 TO 9 /O 1/ K /3 » /S /6 /r /S JS 20 2/ 32 23 ^'^ — Y- ~ — 1 1 V '., 1 \ ', i ; 1 ■ 1 ', / ; I '. - ,' I 1 1 \ - I \ ■ / u r 1 ', / i \ 1 k ' . . .^; 7 1 \ 1 ', / ; / \ f } ; • / ; V r. \ ,•■ 1 / / ; \ / ; / ; ', / / \ ', ,' ; ', r \ 1 s / 1 / 1 \ •_ [ \ y \ li ', A 1 r- ' \ \ ; rl \ ;' f \ . \ ; \ 1 *, 1 , > / ; > > \ 1 \ ' \ f 1 \ » V ,' ' / \ / . 1 \ 'v ^ r- * i' Fig. 5. — Saperda Candida. Diagram illustrating the relative time of emergence of male and female beetles. Based on 261 males and 206 females that issued under natural conditions at French Creek, W. Va., in 1914, 1915, 1917, and 1918. SEASONAL PHENOMENA OF THE HOST TREES AS AN INDEX TO THE TIME OF DEVELOPMENTAL CHANGES OF THE INSECT. Since this borer occurs in North America from southern South Carolina and Texas northward into Canada the calendar dates of its metamorphic changes in different latitudes must vary considerably. There must also be a considerable yearly variation in the dates of these changes in any given locality, due to the early or late advent of spring. In the rearing work with this species it was found that between Demorest, Ga., and Winthrop, Me., there was a difference in time of 28 BULLETIN" 847, U. S. DEPARTMENT OF AGRICULTURE. emergence of the first beetles of about 40 daj^s. On Grand Island, in northern Michigan, the first beetles appeared 75 days after the date of the first appearance in Georgia. In the several years during which beetles were reared at French Creek, TV. Va., there was a variation of 13 days in the dates of the first adults to issue. Calendar dates are therefore of little value in expressing the time when a given metamorphic change of the insect takes place. It was found, however, that the time of certain transformations and activities of the borer may be anticipated or determined very conveniently by observing the definite annual steps in the development of foliage, flowers, and fruit of the apple and other trees upon which the insect lives. The first blossoms to appear on apple follow closely the first activities of the borers in the spring and it is just in advance of apple blossoming time that the first fresh castings thrown from trees by borers may be looked for. Also, the blooming time of apple corresponds quite definitely with the pupal period of the insect. The oviposition time of the beetles begins with and extends somewhat beyond the ripening season of the fruit of the service tree. These rules hold good in a general way for all latitudes and altitudes and for early and late springs. The following field notes arranged in Table X indicate the coin- cidence of these events in a number of different localities : v Table X. — Indicating the correspondence in time of certavn. developmental changes in the roundheaded apple-tree borer and its host trees. Ivocality. Date. Field note. Frenchton, W. Va Weston, W. Va Great Cacapon, W. Va. Kckens, W. Va Winthrop, Me Gadsden, Ala Domorest, Ga Biltmore.N.C French Creek, W.Va., Do Do Munising, Mich French Creek, W. Va. . Do Do Apr. 27,1914 Apr. 28,191-1 May 5, 1914 May 20,1914 Jime 22,1914 Apr. 29,1915 May 1, 1915 May 4,1915 Apr. 20,1916 Apr. 25,1917 Apr. 27,1917 June 20, 1917 Mar. 26,1918 May 15,1918 June 13, 1914 Blossoms of York Imperial and Maiden Blush one-half open. One pupa of S. Candida found. Apple trees In full bloom. More than half the transforming borers have pupated. Apples a little past full bloom. About 25 fresh pupa? of S. Candida collected. Tliree-fourtlis the apple petals off. Maturing borers all in pupal stage. Apple blossoms have been off 2 weeks or more. Pupte of S. Candida still present. A few have issued. Apple petals have been ofl 6 days. One pupa of S. Candida foxmd. Apple trees just past full bloom. All maturing S. Candida in pupal stage. Do. First apple blossoms opened to-day. Half the transforming borers have pupated. First apple blossoms opened Apr. 22. All maturing borers have pupated Apr. 25. Borers in apple throw out first castings a few days in advance of first apple blossoms. First apple blossoms opening. S. Candida found. Apple buds showing first pink. ingfrom l-year-old borers. Last petals falling from apple. Transforming borers all pupie. except one male which has changed to beetle. First fruit of service ripened May 29. First eggs of S. Candida June 4. Fruit of service overripe June 13. Kgg laying of S. Candida at height June 13. About a dozen fresh pupse of Fresh castings first appear- ROinSTDHEADED APPLE-TREE BORER. 29 NATURAL ENEMIES. Possibly no other economic insect of equal importance has had so few natural enemies recorded definitely and specifically as has the roundheaded apple-tree borer. In all the literature upon this borer, there seems to be only one original reference to such an enemy, this being the single instance of the hymenopterous parasite Cenocoelius populator Say, reared about 30 years ago by Riley and Howard (3, p. 59) from borers of this species received from Indiana. Felt and Joutel (6) state that an undetermined carabid larva was found preying on the borers by Walsh and Eiley, and practically all ob- servers have noted that woodpeckers are an important enemy, although in no case is the specific identity of the bird or birds established, so far as the records show. In rearing and handling many thousands of the borers in various* localities the writer has never found any evidence of hymenopter- ous parasites. In two instances undetermined carabid larvse were found devouring yomig borers in West Virginia and another half- grown borer was found that had been killed by a hairworm, sections of the worm being found in the burrow entwined around and within the dead and shriveled body of its host. A large spider was seen to poimce upon and bite in the back a female beetle that had just issued from her exit hole in a tree. In an effort to rescue the beetle the spider was crushed beyond recognition. The beetle died a few hours later from the wound. WOODPECKERS. While the control effect of parasites and predacious insects on this borer is negligible, woodpeckers play an important part in holding it in check. Wherever the writer has collected specimens or made observations in borer-infested localities the work of these birds has always been in evidence. Soon after the borers hatch the wood- peckers begin to find them beneath the tliin covering of bark and thereafter the birds drill for them as long as they are in the tree. In several orchards where counts were made from 50 to 75 per cent of the borers had been destroyed in this way. During October, 1915, 24 young borers were collected and planted in furrows gouged out of the wood beneath loosened tongues of bark on the trunk of an apple tree. A week later, when the tree was revisited for the purpose of putting a wire screen around the trunk to protect the borers from birds, woodpeckers had punctured every tongue of bark and removed the borers from beneath. Not one had escaped. In May of the same year, while pupa? were being collected from an orchard, a total of 11 pupal cells were found and from every one the occupant had been removed by woodpeckers. In another case 30 BULLETIN 847, V. S. DEPARTMENT OF AGRICULTURE. 21 pupal cells were found, 19 of which had been opened b}^ wood- peckers and the insects removed. During the winter of 1915 the writer had standing near his office window a young apple tree in which there were known to be three borers ready to pupate the following spring. The borers had been protected previously by a wire screen but now the screen was removed. On December 21 a male downy woodpecker, D}'yohates fubescens medicmus (Swains), was observed to alight on the base of the trunk and move about alternately tapping the bark and assuming a listening attitude. Presently, with a few vigorous strokes, it drilled tlirough the bark at the point where the future exit hole of a beetle was to have been and at once drew forth and swallowed a large borer. (PI. V, B.) A minute or two later it located a second borer, disposed of it in the same way, and then flew away without further search. Again, in January, 1916, the trunk of a young apple tree known to contain full grown borers was planted in a natural position near the same office window. A few days later a pair of downy woodpeckers came to 'the tree and after a brief search the female was seen to re- move and swallow a borer. A little later the male found and re- moved another. The birds would move about over the trunk tapping lightljT^ wdth their beaks until the quarters of a borer were located. Then with a few sharj) strokes they would j)enetrate to the burrow and remove and devour the insect. The female bird located and removed her specimen through the partly prepared exit hole in less than a minute, but the male drilled industriously for his nearly five minutes, making during the time several openings into the wood which extended in a line over the burrow made by the borer in ascending the trunk to prepare its pupal chamber. Other observations were made which indicate that the hairy woodpecker, Dryobates villosus viUosus (L.), also destroj^s the borers, but this bird was not seen in the act of removing the insects from the tree. METHODS OF CONTROL. Ever since the roundheaded apple-tree borer was first recognized as a serious orchard pest, two principal ways of combating it have been advocated : First, the worming process, in which the borers are removed from their feeding places in the tree by the use of a knife and other tools (PI. IX, B) ; and, second, the covering of that portion of the trunk of the tree where the eggs are most frequently laid with some protective wash, paint, or mechanical device which will act as a barrier against the female beetles during oviposition. Both of these methods are commonly practiced in orchards and have been the lines of most frequent and extensive experimentation by investigators of borer injury and control. In the present studies, modifications of ROUNDHEADED APPLE-TREE BORER. 31 these two methods have received special consideration. Tests were made of the effects of penetrating liquids of an irritating or poison- ous nature when applied to the bark beneath which borers were feeding, of gaseous and poisonous liquids injected into the bur- rows, of sticky substances applied to the trunks of trees for the pur- pose of entangling the adults during their egg-laying activities, and of killing the adults by the use of poisonous sprays. Studies were made also of the distance which female beetles may fly in search of trees in which to oviposit, with the idea of determining the possi- bility of preventing the infestation of orchards by destroying near-by breeding places. These various tests are described in detail below under their various headings, WORMING, The labor of removing borers from trees with a knife and wire is not relished by the majority of orchardists, and yet the difficulties and expense of the task are less than in many other necessary opera- tions in dealing with insect and fungous enemies. Two men, on an average, with an insignificant expenditure for tools and material, should worm 500 trees in a day and obtain as high a percentage of control as ordinarily results fi'om a spraying operation against the codling moth or San Jose scale. Not only does a thorough worming of an orchard rid the trees of the borers present at the particular time but it insures a decreased number of borers for the following one or two years. As is shown on pages 20 to 24, by preventing adult borers from developing within and adjacent to an orchard a reduction may result of about 75 per cent in the number of borers that will attack the orchard the ensuing year. The process of worming is well understood ; the best tools for the purpose being a garden trowel for removing the earth and litter about the trees, a pocketknife with a long, sharp blade, a narrow chisel for securing borers that have penetrated deep into the wood, and a piece of slender wire (PL IX, B) about a foot in length with a sharp hook bent at one end and a tag or bit of conspicuous cloth at- tached to the other end to safeguard the wire against loss. These tools may be carried conveniently in a small fruit basket. The writer found that worming can best be done by two men working together on opposite sides of the tree. With a little practice, one becomes quite adept at locating burrows and hooking the borers from their retreats. After a little skill has been acquired the chisel will have to be used only on rare occasions when deep burrows in the Avood are found to be so crooked that the wire on being inserted will not follow the openings. Worming should be done as soon as possible after the last eggs of the year have hatched, as young larvae usually feed rapidly and 32 BULLETIN 847, U. S. DEPARTMENT OF AGRICULTURE. often injure small trees severely the first season. The proper time for the autunyi worming varies tAvo months or more between the southern and northern limits of the insect's range, and no definite date can be given which applies to all localities. A safe rule is to have the worming job over before the time arrives for gathering the first winter apples. The borers continue to injure the trees during warm weather of late autumn and early winter, often ejecting their castings in the latitude of West Virginia as late as December 1. It is best to prevent all possible injury by getting the worming done previous to the press of ai^ple-picking time. Trees should have a second worming in the spring soon after the blossoming time of the apple, as it is practically impossible to secure all the borers at one examination. Borers are usually easy to locate by their fresh cast- ings soon after apple trees bloom. (PI. VII, B.) At present no cheaper or more effective method of combating this borer is known than that of worming. In order to get best results, however, the work in the orchard must be done thoroughly, and near-by breeding j^laces, such as scattering growths or clumps of apple, wild crab, mountain ash, hawthorn, and service trees elimi- nated either by destroying the trees or by worming. Many orchards are wormed thoroughlj^ ever}'^ year, and just as regularly beetles de- veloping in adjacent trees fly over the fence and provide annually for other generations of borers. For most effective control, there- fore, the worming operation should include not only the orchard but the trees of the locality immediately surrounding the orchard in which borers breed, and the trees should be examined twice annually, first in late summer after the egg-laying season is past and again in the spring after the blossoming time of the apple. WASHES, PAINTS. AND MECHANICAL PROTECTORS. Various materials and devices have in the past been applied to the trunks of trees, either to prevent the female beetles from getting at the bark to oviposit or to kill the borers while feeding in the bark or wood. For preventing oviposition protective coverings, either of a liquid or mechanical nature, have been tested, and, for killing the borers, penetrating poisonous or irritating liquids have been recommended. In the present investigations about 50 kinds of washes, paints, and mechanical devices were tested as to their effectiveness in preventing egg laying and for killing the borers within the trees. Many of these materials were homemade or homemixed and many others were conmiercial products purchased either from the manufacturers or on the market. Nothing in the way of trunk protectors was tested, however, that gave satisfactory results in all cases. Some applica- tions afforded full or a considerable measure of protection against Bui. 847, U. S. Dept. of Agriculture. Plate IX. ROUNDHEADED APPLE-TREE BORER. A, Beetle gnawing bark from apple branch. B, TooLs for use in worming trees. C, Apple branch denuded of bark by beetles. R0T7NDHEADED APPLE-TREE BOEER, "" 33 ovipositing females but could not be applied safely to trees on account of the injury to bark or wood. All degrees of tree injury were ob- tained, consisting of a slight yellowing and dropping of the leaves, checking of growth, roughing and cracking of the bark, rank growth of water sprouts, and killing outright. Some forms of protectors caused the beetles to lay their eggs higher up the trunk than is the custom, the only apparent advantage in their use being that the re- sultant borers were easier to get at in the worming process. Some other materials, such as white-lead paint, gave excellent results in certain cases, and in others where the same material was used in the same way, the female beetles bit through the coat of paint and deposited eggs freely in the bark beneath. In practically all cases, the time and expense required to make and apply protectors of this entire class are greater than those called for in the worming opera- tions, and the results in controlling the borers are less satisfactory. PROTECTORS USED AGAINST OVIPOSITING BEETLES. Since a large proportion of the eggs of this borer are normally deposited within the bark of a limited space just above the ground, it would seem a simple matter to cover or protect in some way that part of the trunk so as to force ovipositing females to go elsewhere to lay their eggs. A considerable number of such protectors were tried over a series of years in a young apple orchard of a thousand trees planted for experiment purposes at Elkins, W. Va. The or- chard was set in rows of 31 trees each, and most of the materials were applied to trees of a single row, leaving the trees of an adjoin- ing row untreated to be used as checks. In every case where paint- like materials were used for more than one year fresh applications were made annually. This was necessary for the reason that the growth of the trees caused all substances to crack and expose areas of the bark. The results of several of these tests are given below in Table XI. Table XI shows that a considerable measure of control was ob- tained by most of the protectors used. None, however, was entirely satisfactory in every respect. In addition to the protecting materials mentioned in Table XI, a large number of others were tested. These included proprietary and commercial products in the form of paints, soaps, tar products, whitewash combinations, viscous substances, nicotine washes, and paper and metal contrivances, all intended to keep the female beetles away from the bark either by offering mechanical barriers or by making approach to the bark so difficult or disagreeable that they would go elsewhere to oviposit. None of these was without objec- tionable qualities, either from the high cost, injury to trees, or lack of effectiveness in keeping out the borers. 34 BULLETIN 847, U. S. DEPARTMENT OF AGRICULTURE. '^ a. & M S a £ fe i S t»c3 a 55 "•ffoJ'o e3 3o t^ OS -^ o CO Tt< ec t-* u3 o o o »o lo OS i-i C^ O r-I 1— I CO t^ CD 00 O 00 "*«^ <£) CO CO C CO CO 1-t §5^ ^ a .^ta ® 03 •c.9 Sfe- Oi t^ (M CO CC Cs Ol CO C^ O CS OS CO lO lO OS O l-l -^ W •'tl 1-H CO l-H l-H ipiOIN iiOCViiO -OOOJ t-^(rO(N'^'^^^^^ PQj^g^j.^^^.^g^l^ Floors and % '"■ Jl III 1 If Wmim^l^^^k walls should be so joined that there is no opportu- nity for accumulations along sides and corners. Well-cemented basements that are light and dry are an aid. Machinery should be placed high enough to allow frequent and thor- ough cleaning beneath it. Where practicable the bottoms of flour conveyors should be of metal and rounded, so as to permit the least amount of flour or meal to accumulate along the side and at the ends. The hoppers of the rolls should be con- structed of cement and in accumulate in inaccessible Fig. 4. — E'.evator and belt brush for cleaning ele- vators infested ty the Mediterranean flour moth. It is made by taking a piece of IJ-inch hoard of same dimensions as elevator cups, fastening bristles to three sides. Side A is fastened to elevator bolt with flat-headed bolts running through board, as shown at BB, the bolts being i or g inch. Bristles on sides CC should be i-inch long, but those at D should be longer, so that a good brushing to outside of elevator may be secured. Such a brush can be made to fit any elevator. (Chittenden.) such manner places. as to allow no flour to USE OF LIME. A liberal use of air-slaked lime in dark corners of damp basements will not only serve as a repellent to insects, but will also tend to destroy some of the objectionable odors and sweeten the air. CARE OF SACKS AND BAGS. New sacks and bags should not be stored in packing rooms or in any place where they become dusted with flour or cereal products, for in this flour dust various mill pests can breed and become established ready to attack stocks packed in the sacks. This is an important point often overlooked. Secondhand sacks should never INSECT CONTROL IN FLOUR MILLS. 9 be taken into the main mill building before they are thoroughly disinfected. It has already been stated that in the majority of cases the spread of moths from mill to mill is traceable directly to the use of untreated secondhand sacks. Such sacks should be stored in a Fig. 5. — View in basement of flour mUl. Note accumulations of webbed flour on floor, on bundles of twine, clinging to brick walls, and hanging from overhead beams. The flour in the foreground on floor is Ijadly webbed. Such conditions are not allowed to exist where control measures are put into successful operation. small, tightly constructed house separated from any part of the mill, and there treated either by heat or fumigation. This may in- volve extra handling but should become a part of the routine of protection against pests. 183545°— 20 2 10 BULLETIN 872, U. S. DEPAKTMENT OF AGRICULTURE. Fig. 6.— Portion of side of very poorly kept flour mill. Note accumulations of flour, much of which is webbed by the flour moth, on side walls, timbers, and ladder. Such neglect breeds trouble. INSECT CONTROL IN FLOUR MILLS. 11 CLEANLINESS IN WAREHOUSE ROOMS. Warehouse rooms, as well as the mill proper, should be kept as clean as possible. After each movement of stocks the floor and walls should be cleaned to prevent insects present from holding over to attack a new stock. Neglect of this simple precautionary measure has resulted in a more rapid infestation of warehouse stocks than is generally appreciated. NATURAL CONTROL BY PARASITES. In many mills the Mediterranean flour moth is attacked by insect parasites, TIdbrohracon hebetor (Say) and Omorga frumentaria (Rond.), but in no case are these of sufficient importance to warrant a miller depending entirel}' upon them, although at times they may prove a valuable control factor. They are not a dependable factor in any part of the United States. In a successful fight against the flour moth, control by parasites should not be relied upon, ARTIFICIAL CONTROL MEASURES. Since the spread of the Mediterranean flour moth from Europe to the milling centers of the United States experimental work has been conducted to determine the most satisfactory methods of con- trol. _From the various processes advocated from time to time, such as fumigation with sulphur, carbon disulj^hid, tobacco fumes, for- maldehyde gas, etc., there have finall}^ emerged two control measures that have now proved their value. These are fumigation with hydro- cyanic-acid gas and control by heat. Where these can be intelli- gently and thoroughly applied results can be guaranteed. Heat is unquestionably the cheapest and simplest agent of control now knf)wn. Millers have had varying results in attempts to control the moth with freezing temperatures. HYDROCYANIC-ACID GAS TREATMENT. Fumigation with hydrocyanic-acid gas, while effective, is a dis- agreeable method of control. The gas generated is deadly to man, as well as to insects, and precautions must be taken to safeguard the lives of laborers and dwellers in closely associated buildings. It is as- sumed for the purposes of this bulletin that no milling concern will attempt fumigation with hydrocyanic-acid gas without calling in the services of an expert familiar with the process.^ It may be added that where it can be used, hydrocyanic-acid gas fumigation is per- fectly safe when conducted under the immediate supervision of a ~ For more complete information regarding the use of this gas, write to the U. S. De- partment of Agriculture for publications. 12 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. properly informed and careful person. Success in the use of hydro- cyanic-acid gas is in proportion to the tightness of the mill to be fumigated. Many old loosely constructed mills can not be fu- migated successfully, even by an expert fumigator, because they will not hold the gas sufficiently long. In such mills one can not hope to do more than reduce the moth temporarily by any one fumi- gation. In tight structures two thorough fumigations applied dur- ing warm Aveather Avithin a short time of each other, as discussed on page 27, have been known to exterminate the moth. In rare cases one fumigation has exterminated the moth. I'RECAUTIONAEY INIeASURES. Although it has been stated above that no mill concern should attempt fumigation without the assistance of a (^ h e m i s t , entomologist, or other person Avell informed I'egarding the process, -atten- tion should be called to sev- oral precautions that must be taken to guard against fa- talities arising from inex- cusable carelessness and igno- rance. Very few deaths have occurred as a result of fu- migation with hydrocyanic- acid gas as a mill fumigant, and those that have oc- curred have been due to criminal neglect on the part of those responsible for the fumiga- tion. AMiere mills are in buildings well isolated from other buildings, owners need not consider neighbors in their plan for fumigation, but when the fumigation is to be conducted in one of a series of buildings, such as occur in city blocks, owners of property immediately adjoin- ing must be informed of the intention to fumigate and arrangements made with them whereby they will be ready to vacate their stores or offices should the fumes penetrate the intervening walls. The odor of hydrocyanic-acid gas is easily detected, and the person in charge of Via. 7. — Wooiifii oil barrel, well soaked with water previous to use in fumigation, set in galvanized iron waslitub partially filled with water to which have been added sev- eral handfuls of sal soda. INSECT CONTROL IN FLOUR MILLS. 13 fumigation should advise regarding the need to vacate. Persons should be advised not to remain in rooms closely associated with rooms being fumigated if the odor of the gas can be detected. In numerous fumigations persons have remained at their work even when the odor of the gas was very pronounced, and workmen have been known to carry on trucking operations in warehouses where the odor of the residual gas after fumigation was so strong as to give the writer of this bulletin unpleasant feelings. Rooms adjoining those being fumi- gated should be kept well ventilated, but even then, to remain in them when the odor of the gas can be detected may lead to fatalities directly chargeable to the fumigator. To attempt fumigation without the approval and cooperation of those controlling the occupants of ad- joining buildings may lead to fatalities directly chargeable to criminal neglect on the part of the person conducting the fumigation. Guards should be placed at entrances that can not be closed by means of reliable locks to thoughtless intruders. Only trusted men should be employed to assist in any process of fumigation, and this holds particularly true of the person or persons left on guard. One man, at least, is known to have been killed because ho was not in- formed of the fumigation, returned for night work, and entered through an unlocked door left temporarily unguarded by a sleeping guard. A genuine responsibility rests upon a person conducting fumiga- tion and he should discharge his duties with all seriousness. He should choose as helpers only the most intelligent and trustworthy employees. Preparation of Mill for Fumigation. Preliminary cleaning. — In fumigating a mill the greatest loss of time is caused by the need for preliminary cleaning; the fumiga- tion can be done between Saturday night and early Monday morn- ing or within any period of 24 hours' duration. The reason for thorough cleaning and destruction of flour and accumulations from various parts of the mill is primarily to leave in the mill at the time of fumigation no masses of stocks which the gas must penetrate to reach all the insects. Hydrocyanic-acid 'gas, deadly as it is, can not be depended upon to reach and kill all moths, pupae, eggs, and larvse if these are buried beneath several inches of well-packed flour, such as often occurs in dead spaces of machines, in floor cracks, basements, etc. The removal and destruction by burning of such accumulations which, as has been stated already, are favored by pests as feeding and breeding places, destroy large numbers of insects, as well as leave the mill in a condition to be fumigated to best advantage. Make mill as tight as possible. — The mill should be made as tight as possible, so that the gas generated can not escape before it has had a chance to kill the flour moth. Broken windows, ventilators, 14 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. and cracks about window frames can be made tight by pasting strips of paper over them. Even an occasional open window and doorAvays, stairways, and elevator shafts can be fairly well sealed by the use of heavy paper applied in several thicknesses by paper hangers. All such openings must be closed before fumigation is attempted or fail- ure will result. As the gas has a tendency to rise, each floor should be treated as a unit by closing such openings as those formed by ele- vator shafts, stairways, etc. The small openings through which INSECT CONTROL IN FLOUR MILLS, 15 belting passes and similar openings in the floor can be sufficiently well closed by crowding pieces of sacking firmly into them. Chemicals Necessary for Fumigation. The chemicals necessary for the generation of hydrocyanic-iicid gas are potassium or sodium cyanid,^ sulphuric acid, and water. ' Cyanid. — Either potassium cyanid or sodium cyanid can be used in fumigation, but trade conditions have made sodium cyanid more available for use and, as it is just as good and slightly cheaper, it is recommended. For a full discussion of the relative values of potas- sium and sodium cyanid for fumigating purposes, see Bulletin 90 of the Bureau of Entomology. Sodium cyanid for fumigation pur- poses should be 96 to 99 per cent guaranteed purity. The volume of gas liberated is in direct proportion to the purity of the cyanid, hence too much stress can not be placed on purchasing high-grade cyanid. Protection can be had by purchasing a cyanid guaranteed under the Federal insecticide act; the analysis on the label should show that the cyanid is at least as pure as follows : indium ctinnid {XaCN), 96 to 99 per cent, an/ilysis. Per cent. Cyanogen (CN), not le. s than .51.3 Sodium (Na), not more than 4.3.7 _. Inert substances, not more than 4. Of; j Chlorlds, not more than . 1.4 High-grade cyanid can be purchased in tin cans containing 50, 100, or 200 pounds. For ease in handling and preparing doses, it is con- venient to purchase brands of cyanid in which the material is divided into lumps weighing approximately 1 ounce each. Cyanid decomp<(j)ses slowdy when exposed to the air, hence such cyanid as remains unused after a fumigation should be protected by being placed in an air- tight receptacle. Where this can not be done, the filling in of 'the original container with firmly packed sacking will greatly retard decomposition. Sulphuric acid. — Commercial sulphuric acid (H2SO4) 92 to 94 per cent pure (66° Baume), free from nitric acid, arsenic, lead, and zinc, should be used. It may be made from pyrites or pure brim- stone, so long as impurities are eliminated. Acid is usually pur- chased in iron drums containing 800 to 1,500 and 2,000 pounds, though when used in smaller quantities it can be had in glass carboys of about 100 pounds capacity. Pure acid is colorless and about twice as heavy as water; its specific gravity is 1.83. When stored in iron drums it ^ Recent developments in liquid hydrocyanic acid indicate that sooner or later its use will replace this long established method of generating gas by the mixing of chemicals in containers. 16 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE, frequently has a milky white color, especially at the bottom of the drum, due to the presence of sulphate of iron resulting from the Fig. 9. — Second floor of a flour mill, showing the proper arrangement of the generators and the bags of cyanid for hydrocyanic-acid gas fumigation where the bags are to be dropped by hand. (Dean.) action of the acid upon iron either before or after the acid is placed in the drum. This sediment does not affect the value of the acid INSECT CONTROL IN FLOUR MILLS. 17 unless present in excessive quantities. In estimating the amount of acid required it is sometimes convenient to know that 1^ pints of acid Fig. 10. — Third floor of a flour mill, showing the proper arraugement of the generators and the bags of cyanid for hydrocyanic-acid gas fumigation where the bags are to be dropped by hand. (Dean.) by liquid measure is required for each pound of cyanid when the following formula is used. 183545°— 20 3 18 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. Pboportion of Chemicals. Hydrocyanic-acid gas is generated by mixing in proper proportion sodium cyanid, sulphuric acid, and water. These should be com- bined as follows : 1 ounce, by weight, of sodium cyanid. li ounces, liquid measure, of sulpliuric acid. 2 ounces, liquid measure, of water. This is known as the l-l|-2 formula. Slightly varying formulas have been recommended, but this l-l|-2 formula yields good results. Equipment for Generating Gas. The equipment for generating gas may be verj^ simple and should include the following : Containers for generating the gas. — For mill fumigation, 4-gallon earthenw^are stone crocks, well glazed, such as may be purchased at any hardware store, have proved very satisfactory. Four-gallon crocks (see figs. 8 to 14) will receive a charge containing a maxi- mum of 4 pounds of cyanid. Fumigation does not injure them for other purposes provided they are properly cleaned. Where very large spaces are to be fumigated strong water-tight wooden barrels have frequently been used. Untreated oil barrels in good condition have served for several fumigations without marked deterioration, where care is taken to add the acid slowly and to rinse thoroughly after using. A 50-gallon barrel will accommodate with ease a charge of 20 to 30 pounds. Containers holding larger quantities of solution are more difficult to remove and empty than smaller ones, and this is an important item. Good barrels sufficiently small to fit easily into an ordinary galvanized-iron washtub (fig. 7) and receive charges of 15 to 30 pounds of cyanid are very satisfactory when buildings are so arranged that they can be easily removed and emptied. If the tubs are partly filled with water to which have been added sev- eral handfuls of ordinary salsoda, there is no danger that the acid will spill out onto the floor during the chemical reaction which pro- duces the gas, and the handles of the tub make it easy for two men to remove the tub and barrel at the same time after fimiigation. Measuring and carrying equipment. — Scales are needed for weigh- ing cyanid. Glass graduates of 16 or 32 ounce capacity should be used for measuring acid and water. One or two gallon granite ware cups and pitchers are very convenient and safe for carrying acid. Ordinary granite ware buckets in good condition can be used very satisfactorily in carrying acid from carboys to generators, though these should be carefully washed immediately after use and not allowed to stand with acid in them. Containers of tin must not be used. INSECT CONTROL IN FLOUR MILLS. 19 Sachs. — Sacks for holding the individual charges of cyanid are needed. \Vlien 4-gallon crocks are used common manila paper sacks, sizes 8 and 10, such as may be had of any grocer, are the best Fig. 11. — Fourth floor of a flour mill, showing the proper arrangemeut of the generators and the bags of cyanid for hydrocyanic-acid gas fumigation where the bags are to be dropped by hand. (Dean.) in which to place the cyanid. The cyanid is placed in the smaller sack, after which the smaller sack is slipped Avithin the larger one. Sacks of heavy or heavily glazed paper should not b.e used, for 20 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. they retard the action of the acid upon the cyanid. Where larger doses must be carried old bran sacks may be used for the cyanid. Protection against hums and charring. — Persons engaged in fumi- gation according to the method described in this bulletin invariably receive slight injuries to their clothing in the form of acid burns due to spattering and dripping of acid. Only the oldest clothing should be worn. A saturated solution of common washinjr soda crystals (sodium carbonate) should be at hand ready to apply to hands, face, or floor to neutralize acid spatterings that may occur. When acid is purchased in iron drums a sufficient number of 5 to 10 gallon glass carboys in the usual wooden frames to hold all acid Fig. 12. — Diagram showing " stringing method " used in generating hydrocyanic-acid gas. The number of generators connected with the main cord may vary to fit the need of individual cases. (Chittenden.) needed for the fiunigation, should be provided to hasten the meas- uring of the acid. To guard against injury to the mill flooring, the acid should be measured outside the mill; or if this work is done within the mill old sacks and sawdust should be used liberally to prevent any acid spilled during the work from damaging the floor. Provision for ventilating rooms after fumigation. — Care should be taken to provide for tlie ventilation of rooms or buildings after fumi- gation. It is dangerous to enter rooms to open windows from within although experienced fumigators often take such risks. Ingenuity can be depended upon to discover a way to open windows, transoms, or skylights from the outside by means of cords or wires. INSECT CONTROL IN FLOUR MILLS. 21 Estimating Spaces. After a decision to fumigate has been made, the amount of space to be fumigated must be determined by securing the inside measure- ments of the length, width, and height of each room. No deductions should be made for machinery, bins, etc. Amount of Chemicals Needed. Once the cubic contents have been determined the amount of chemi- cals needed can be calculated easily. In the treatment of single tight rooms 1 ounce of cyanid is recommended for each 100 cubic feet of space. Where a reasonably tight mill has 4 or 5 floors the follow- ing standard, suggested by Dean, should be followed, as the gas is light and has a tendency to concentrate on the upper floors. Use 1 pound of sodium cyanid to each 1,000 cubic feet of space in the base- ment, to each 1,200 cubic feet of space on the first floor, to each' 1,300 cubic feet of space on the second floor, to each 1,500 cubic feet of space on the third floor, and to each 1,600 cubic feet of space on the fourth and fifth floors. When buildings are rather loosely con- structed, a larger amount of cyanid will be needed, but no directions can be given that will be as useful in determining the amount neces- sary to offset the looseness of individual mills as one or two experi- mental fumigations. According to the 1-1^-2 formula recommended on page 18, for each 100 pounds (1,600 ounces) by weight of cyanid required there will be needed also 2,400 fluid ounces of sulphuric acid. Number of Generators Required. If 4-gallon crocks are used for generating the gas, one crock will be needed for each 4 pounds of cyanid required, or for each 6,400 cubic feet of space, when 1 ounce of cyanid is used to each 100 cubic feet of space. ^Vhere larger generators are used, as discussed on page 18, determine the amount of cyanid to be used in each container and use it to divide the total cyanid required for the building or room to secure the number of generators needed. It is always well not to overtax the capacity of a generator. Placing of Generators. In placing the generators or jars, care must be taken to have them sufficiently distant from machinery, belting, sacks, walls, etc., so that there may be no danger of injury from spattering during gas genera- tion. The reader is directed to examine carefully figures 8 to 11 for illustrations of how best to arrange generators. In using crocks it is always best to place beneath each crock some old sacking or papers, as not infrequently there is a slight spattering of the floor 22 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. close to the crocks during the chemical action which forms the gas. This is particularlj' true when a maximum charge for any container is used. Orukr of Assembling Chemicals for Final Generation of Gas. Having decided upon the number of generators for each floor and the amount of charge for each, the generators should be well dis- tributed. The proper amount of water should then be measured out INSECT CONTROL. IN FLOUR MILLS. 23 and poured into each container. The acid should always be added to the water and never vice versa. The mixing of water and sul- phuric acid generates heat; hence care should be taken to pour the acid slowly into the water to prevent too rapid a rise in temperature, which might cause a cracking of the crocks. As the gas is formed best when the mixture of acid and water is warm, the acid should not be added until just before the fumigation is started. As soon as the acid is poured into the generators the cyanid, which has :ilready been " weighed out into proper amounts and placed in sacks, should be dis- tributed, one sack by each generator, as illustrated in figures 8 to 11. Dropping of the Cyanid. By this time everything should be in readiness for the generation of the gas. The building has been tightly closed, except for the doors or windows through which those who drop the cyanid will leave the building. Provision has been made for ventilating after fumi- gation (p. 20), the building has been cleaned (p. 13), made tight (p. 13), and manholes, slide doors, etc., in spouts, elevator legs, etc., have been opened, and all persons have been accounted for. those not assigned to dropping cyanid having left the building. Starting on the top floor, begin dropping the cj'anid carefully, though quickl}', into the generators, starting at the jars most distant from the exit and working rapidly toward the exit. This requires quick, cool- headed action. If a charge of cyanid is overlooked never go back to drop it ; let it go — your safety comes first. When the last charge has been dropped leave the floor, closing the door behind, and repeat the operation on each successive lower floor. AVhen the person in charge feels that it is not wise to attempt to drop all charges of cyanid by hand because of difficulties in the way of retreat the stringing method should he employed. Novices often resort to this method, though experienced fumigators do not, except in difficult situations. Where floors are so crowded with machinery that a hasty exit can not be made, or in dropping charges in tlie basement where one must run upstairs to the first floor before reach- ing the exterior the use of the stringing method will eliminate danger. This method consists in passing strong cord through con- veniently arranged screw-eyes, each string finally passing from a central point at the exit to and through a screw-eye in the ceiling, so that it will hang directly over the spot where the generator is to be placed. (See figs. 12, 13, and 14, J^.) The strings should be so arranged that the}' may all be lowered into the jars by releasing one main string. This may be done by carr^nng the strings leading to each jar through screw-eyes in the ceiling or wall to one stout cord at the exit, as illustrated in figure 12. After the strings have 24 BULLETIN 872., U. S. DEPARTMENT OF AGRICULTURE. FrCI. GENERATOR TOO CLOSE FIG.2. GENERATOR TOO CLOSE TO BAGS OF FLOUR. TO BELTS AND BAG TOO LOW. I FrG.3. GENERATOR TOO CLOSE FIG.4. CORRECT PLACING OF TO BELTS AND BAG TOO HIGH. GENERATOR AND BAG. Fig. 14. — The correct and incorrect placing of generators and bags of cyanid for hydrocyanic-acid gas fumigation. (Dean.) INSECT CONTROL IN FLOUR MILLS. 25 been arranged, the generators should be placed and the water and acid added. Each generator should then be moved at least 2 feet away from where the cyanid will be suspended. Men should then go through the mill, tying firmly the sacks of cyanid to the strings, so that they will hang suspended about 2 or 3 inches above the tops of the generators. This done, the generators are placed beneath the suspended sacks. All persons should then leave the building. One man can now pass from the top to the lower floors lowering the cyanid into the generators by the simple manipulation of the central cord at the exit of each floor. The basement charge should, of course, be lowered last, and preferably by means of a cord running to the outside mill door or to the outside through some slight opening about a window. Closing Buildings and Guarding Doors. After the last charge has been set off and the last person has left the building, the door or doors used as exits should be locked. If such doors do not fit tightly, it will pay to paste strips of paper about their edges. Guards should be placed about the building in such a manner as to prevent persons entering the building. Guards should not stand where they can smell any amount of escaping gas. There should be several guards who keep in touch with one another, espe- cially during the first one or two houi-s after the generation of gas. Time to Fumigate and Length of Exposure. Fumigate only during calm weather. During high winds the gas is carried to one side of the room or building and, unless the building is very tight, it is dissipated more quicld3\ Experiments have proved that insects are not active at temperatures lower than 50° to 60° E. and are not then so easil}^ killed by the gas. Best results follow fumi- gations when the temperUture is 70° F. or above. Everything for fumigation should be in readiness before dark. When possible the best time to fumigate is from Saturday afternoon to Sunday evening or Monday morning. At this time fewer people are about. The buildings should be allowed to remain closed for fumigation not less tiian 12 to 18 hours, though exposures of from 24 to 36 hours, when the building is tight and time is not an important factor, may effect a better killing of pests. Ventilation. Several hours should elapse after a mill has been opened for venti- lation before laborers are allowed to enter for regular work. The doors and the windows should be opened from without, according to provision made before fumigation. In opening windows and doors, open first those on the side of the building opposite to the 26 BULLETIN 872, U. S. DEPARTMENT OF AGRlCULTUEE. direction from which the wind is coming. After the preliminary ventilation has been in progress for one or two hours, it is safe for an operator to enter and open all windows and doors, but he should not remain in the building until it has been thoroughly aired. Specific directions as to the length of time for ventilation can not be given to meet all cases. Much depends upon the movement of air currents, the humidity of the air, and the rate of gas leaking from the building during the hours of fumigation. To be absolutely safe, buildings should be ventilated until there is no odor of gas before persons are allowed to resume their work. Men accustomed to fumi- gate soon acquire from experience knowledge which governs their action in remaining in buildings during the period of ventilation. Pbocedure Aftee Ventilation. After the mill has been ventilated thoroughly an operator should tour the building and make certain that no charge of cyanid was overlooked in the dropping or that no charge failed to be lowered into the generators if the stringing method was used. Any such charges should be removed from the building and placed with the stock of unused cyanid. Laborers should now be sent through the building to remove the generators, diunping their contents into a sewer or into a hole dug in the ground and giving the generators a thorough w^ashing. If tlie chemical action has been complete, the residue in the jars after fumigation may not be particularly poison- ous, yet it is acid, and will burn clothing and skin, and should be handled with care by reliable men. Men should be warned against deliberately inhaling any fumes that may be given off by the residue, either while they are removing the generators or emptying them, (xenerators that begin to bubble when disturbed should be avoided until the bubbling has ceased. When the residue has been absorbed, after it has been poured in the hole in the ground, the soil should be replaced. Efficiency of P^jiigation. Experiments prove that in an air-tight chamber hydrocyanic-acid gas will penetrate in killing concentrations to a depth of 3 inches, but in mill fumigation the gas does not penetrate flour and mill products much beyond 1 inch, and often, particularly near the floor, not even that fai-. Not all mill pests are equally affected by the gas. In a mill infested by the Mediterranean flour moth hydrocyanic-acid gas is a very effective treatment. All stages of the moth, including the egg, if not covered with more than 1 inch of flour, are killed, but upon the several stages of the flour beetles, " bran bugs," and the cadelle the gas treatment is of less value. Many of these pests hide in places INSECT CONTROL IN FLOUR MILLS. 27 inaccessible to the gas and escape treatment only to restock the mill later and offset the large numbers of their species killed during the fumigation. Frequency of Fumigation. Most millers who fumigate with hydrocyanic-acid gas apply the treatment once a year. Certain millers fumigate in the spring, give a second fumigation during midsummer, and a third during early falL A more effective plan is to give two fumigations in July or August within three or four weeks of each other. Of the different stages of the moth, the egg stage is the most difficult to kill, and a certain number of eggs usually escape the first fumigation. If, then, a second fumigation is applied three to four weeks after the first, it will catch the worms that hatch from the eggs unaffected by the first fumigation, when they are in a stage easily killed by the gas ; at that time they have not had an opportunity to develop into adult moths and lay more eggs. Fumigation is advocated only for the warm sum- mer months. Effects of Hydrocyanic-Acid Gas Upon Flour. Experiments conducted at the Kansas Agricultural Experiment Station, results of which are published in Bulletin Xo. 178 of that station, included fumigations of four grades of soft winter-wheat flour, consisting of a patent, a straight, a clear, and a low grade, and of three grades of hard winter wheat flour, consisting of a patent, a straight, and a low grade. Samples were fumigated at the maxi- mum strength recommended for Hour-mill fumigation, viz, 1 pound of cyanid to each 1,000 cubic feet of space. Flour was exposed to a temperature of 90° F in a tight chamber kept at constant temperature for a period of 12 hours. Dean, of Kansas, concludes: It is only in the careful raeasureiueuts employed in the test that any differ- ence between the fumigated and the unfumigated flour is apparent at all. The only notable difference appears in the maximum volume of the dough in the test made immediately after fumigation, but not after thirty days. The finished loaf shows no deleterious effect from fumigation in any of the tests. CONTROL BY HEAT. It has been stated that fumigation with hydrocyanic-acid gas, effective as it is in controlling the Mediterranean flour moth in mills, does not control so satisfactorily the several stages of the flour beetles {Tribolium confusum Duv. and T. ferrugineum Fab.), the "bran bug" {Laemophloeus minutus Oliv.), the cadelle {Tenebroides vuiu- ritanicus L.), and the sawtoothed grain beetle {Sllvanus surinam- ensis L.) These insects are naturally more resistant to the effect of gas, and secrete themselves in large numbers in cracks and accumu- 28 BULLETIN 872, V. S. DEPARTMENT OF AGRICULTURE. lations of fine stocks inacessible to any gas. The flour beetles and J> the cadelle, the larval stages of which cause much trouble in flour, " are in practically every flour mill in the country, except where reme- dial measures are successfull}^ practiced. Inspection of ports in this country and in Europe through which the flour from many American mills is handled, either for domestic or for foreign trade, indicates that these insects just mentioned, which can not be con- trolled so satisfactorily by fumigation, are causing serious trouble, and that the large majority of the infestations originate at the mills. I"iG. l."i. — Photograph of elevator legs after application of the heat method of control. Note the large number of flour beetles M-hich, on feeling the heat, h.ive crawled out of the cracks of the wooden elevator legs and died upon the mill floor. The most practical and effective method known to control all classes of mill-infesting insects is the application of high temperatures. (Fig. 15.) Heat has been recognized as a control agent for many years in this country and in Europe. But its effectiveness as a control measure in flour mills was first demonstrated by professional experimental inves- tigation by Prof. George A. Dean in several Kansas mills during the period 1910-1913.* So successful were Dean's demonstrations that the heat method has been tested by the Federal Bureau of Entomology and by several State entomologists with the result that mills in * DeaNj Geoege a., mill and stoked-grain insects. 189, p. 139-236, 56 figs., 6 pi., July, 1913. Kans. State Exp. Sta. Bui. INSECT CONTROL IN FLOUR MILLS, 29 Kansas, Ohio, Nebraska, Illinois, Indiana, Iowa, southern Canada, Virginia, and elsewhere are equipped for and are using the heat method. The result of all this work is that to-day the heat method is recognized as the most effective, practical, and inexpensive of all treatments, and has the added advantage of being absolutely safe. The greatest exj)ense is the original installation of radiating surface. Yet this expense is offset by the greater cheapness of the heat method as compared with any other effective control measure. Goodwin,* an experimenter in Ohio, claimed in 1912 that for a mill of average size, forced to apply remedial measures, the cost of steam piping necessary to obtain killing temperatures was offset within five years by the saving due to the cheapness of the heat method. Heat Applied ]\Iost Effectively in Summer. It is advised that heat be applied during the sunmier months in order that the normally high temperatures then obtaining may be turned to the advantage of the miller. It is much easier to raise the temperature of a mill to the killing point when the outdoor tem- perature is 85° F. or above. It takes less fuel and less radiation sur- face and saves time. Take advantage of summer heat. Heat only in quiet weather ; it is impossible to get uniformly good results dur- ing windy weather. Degree of Heat Necessary. A temperature of 118° F. to 125° F. in all parts of the mill is suffi- cient to kill mill pests. To obtain these killing temperatures in the least exposed parts, it is necessary to heat other portions to a still higher temperature unless greater care than usual is taken to dis- tribute radiation surfaces according to the situation. For tempera- tures secured in various parts of mills heated experimentally with equipments already installed, see data in Tables I to XII. Use of Steam Heat. Steam heat is the most satisfactory^ heat for raising mill tempera- tures to the killing point. A pressure of 25 to 50 pounds is recom- mended. Amount of Radiation Required. The number of square feet of radiation required to heat a given number of cubic feet of space depends upon the number of doors and windows in a building and upon the general tightness of the structure. Because these factors vary so greatly in mills, a definite recommen- dation for amount of radiation surface can not be given. The engi- * Goodwin, W. H., flour mill fumigation. Bui. Ohio Agr. Exp. Sta., no. 234, p. 184, January, 1912. 30 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. neer of each establishment, aided by one or two experimental heat- ings, will be able to increase or reduce the radiation surface in various parts of the mill until provision is made for securing sufficiently high temperatures. One square foot of radiation is usually sufficient to heat from 50 to 100 cubic feet of space. A mill with sufficient radia- tion to heat it during winter to 70° F. without the heat of the running machinery can be heated readily in summer to 120° F. or 125° F. Dean suggests for a five-story building 1 square foot of radiation to each 50, 60, 75, 90, and 110 cubic feet of space for the first, second, third, fourth, and fifth floors, respectively ; and in a mill of four floors, 1 square foot of radiation to each 50, 60, 75, and 100 cubic feet of space for the first, second, third, and fourth floors, respectively. The amount of radiation surface per linear foot and the linear feet of pipe per square foot of radiation surface for pipes of varying sizes are as follows : Size of pipe, in inches. 1. 2i Linear feet Radiation of pipe per surface per square foot linejar foot. of radiation surface. 0.346 2.9 .434 2.3 .494 2.0 .622 1.6 .753 1.3 If steam pipe is used for radiation the l^-inch or 14-inch size is recommended as most practical. Heat does not Injure Equipment. Heating mills as recommended for the control of mill pests will not injure the mill structure or equipment. In the many heatings on record no injury has occurred to belting or machinery, or, by checking, to elevator legs, woodwork of bolters, and purifiers. Mills have been heated to as high as 150° F. for 30 hours without injury to any part. Objection made that insurance companies will not permit the heat method of control is without foundation. Mr. William Reed, as secre- tary of the Mutual Fire Protection Bureau, representing eight of the principal millers' insurance companies, in a notice to all policy holders stated, "We propose to advocate the heating system for effective fumigation against the Mediterranean flour moth, weevil, and all other mill and srrain infesting in.sects." Flour not Affected by Heat IVIethod. Results of baking tests made of patent hard-wheat flour, a low- grade hard-wheat flour, and a pancake flour have proved conclu- INSECT CONTROL IN FLOUR MILLS. 31 sively that the heat method, even at several deo;rees of temperature higher than recommended for mill treatment, has no injurious effect upon the baking qualities of the flour. A low hard-grade wheat flour was subjected to a temperature of 140° F. for nine hours on three successive occasions (the second and third heatings occurring two and six weeks after the first heating), and a pancake flour was subjected to a temperature of 130° F. for 48 hours without injurious results. Effect of Heating upon Mill Humidity. Heating to kill mill insects greatly reduces the liumidity in the mill. Insects die more quickly in a dry heat than in a moist heat. The relative humidity on the second floor of a mill at 10 a. m., when the heat was turned on, was 93 per cent; by 12 m. it had fallen to 40 per cent, and by 5.30 p. m. to 27 per cent. In a second mill the relative humidity at 6 a. m., when the heat was turned on, was recorded by a hygrograph as 72 per cent ; during the first few hours following there was a rapid decrease to less than 40 per cent, and during the afternoon and throughout the night there was a still further decrease to 12 per cent. Important Points to be Considered in the Successful' Heating of a Mill. 1. Steam pipes should be located near floor and arranged to give equal distribution of heat. 2. Provide water trap for drawing off accumulation of water in pipes. 3. Lower floors and floors with heavy machinery should have more radiating surface in proportion to cubic feet of space than upper floors and floors with light machinery. 4. Mill will heat more rapidly with a steam pressure of 25 to 50 pounds. 5. To take advantage of heat in machinery begin heating mill immediately after shutdown. 6. Stairways and elevator shafts should be closed, so as to make separate units of each floor. 7. Thermometers should be placed at different points on each floor that temperatures may be readily ascertained, 8. Time must be taken to reach desired temperatures. ' 9. A temperature of 118° F. to 125° F. is sufficient for any part of mill. 10. This temperature should be maintained for several hours to allow heat to penetrate all infested parts. 11. Do not attempt to heat on a windy, cold, or rainy day. 32 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. 12. By beginning the heat process directly after shutting down mill Saturday afternoon and continuing until Sunday night or early Monday, there is no loss in producing hours. EXPERIMENTAL DATA ON HEAT FUMIGATION. The following temperatures and results were secured in experi- mental work by Dean (mills 1 to 4) and by Goodwin^ (mill 5). Mill No. 1. Brick building;; day (.July 7-S, 1912) calm, partly cloudy; outdoor maximum temperature 91° F., outdoor minimum temperature 73° F. Heating system, steam pipes along walls, except in space beneath first floor where radiators are used. Steam pressure of about 20 pounds maintained during experiment. Ca- pacity of mills, 600 barrels. ON FIRST AND SECOND FLOORS. Location and reading of ihcrmomctcrx. — The thermometers were located on these floors as follows : No. 1, in 2 inches of flour in elevator boot on floor, 8 feet beneath steam pipes ; No. 2, hanging in middle of room. .5 feet high, VS feet from steam pipes ; No. 3, in 2 inches of flour in elevator boot on floor, 12 feet from steam pipes; No. 4, hanging in the open, 4 feet high, 15 feet from steam pipes ; No. 5. between the rolls in a roll, 11 feet from steam pipes ; No. 6, hang- ing in the open, 6 feet high, near roll machinery. Table I. — Data shoiving rise in temperatures on first and second floors. First floor: Capacity, 28,728 cubic feet; radiation, 525 square feet. Second floor: Capacity. 28,728 cubic feet; radiation, 560 square feet. Time of day. Thermc meter- No. 1. No. 2. No. 3. No. 4. No. 5. No. 6. July 7 1912: 10.30 a. m 0 p 83 83 84 88 90 93 94 95 97 98 100 106 106 106 107 108 0 p 90 94 98 104 106 107 110 110 113 115 122 124 125 o p 86 88 91 96 100 102 103 107 108 108 116 117 118 0 p 98 105 110 117 120 123 125 127 129 130 140 142 144 144 147 146 0 p 97 97 98 100 102 104 105 108 109 111 122 125 126 127 129 131 o p 99 11.30 a. m 12.30 p. m 2.80 p. m 3.30 p. m 4.30 p. m 106 111 118 121 123 5.30 p.m.. 125 7 p. m 127 8. 30 p. m 129 9.30 p. m 129 July 8, 1912: 9 a. m 140 11 a. m 142 12m 144 2p. m 128 118 129 120 144 4 p. m - 147 5 30 p. m 129 121 145 ON THIRD AND FOURTH FLOORS. Tiocation and reading of thermometers. — Thermometers were located on these floors as follows: No. 1, hanging in the open, .5 feet high, 15 feet from steam »0p. cit INSECT CONTROL IN FLOUR MILLS. 33 pipes ; No. 2, in flour in a conveyor near floor. 12 feet from steam pipes ; No. 3, in flour In conveyor, 6 feet higli, 15 feet from steam pipes ; No. 4, lianging in the open, 5 feet high, 12 feet from steam pipes; No. 5, in flour, in a conveyor near floor, 12 feet from steam pipes; No. 6, in flour, in a conveyor near floor, 12 feet from steam pipes. Table II. — Data showing rise in temperatures on third and fourth floors. Time of day. Third floor: Capacity, 31,122 cubic feet; radiation, 460 square feet. Fourth floor: Capacity, 43,092 cubic feet; radiation, 400 square feet. Thermometer. No. 1. No. 2. No. 3. No. 4. No. 5. °F. °F. "F. "F. S5 89 96 88 SS 91 100 90 91 95 105 92 100 102 114 99 102 103 116 101 105 107 lis 103 107 108 119 106 111 111 120 109 113 113 121 111 114 114 122 112 126 125 127 lis 127 125 129 119 127 126 132 120 128 128 133 121 128 131 138 124 129 131 138 124 No. 6. July 7, 1912: 10.30 a. m 11.30 a. m 12.30 p. m 2.30 p.m. 3.30 p.m. 4.30 p. m. 5.30 p.m. 7p. m 8.30 p.m. 9.30 p.m. July 8, 1912: 9 a. m 11 a. m 12m 2p. m 4p. m 5.30 p.m. 95 100 105 114 116 119 122 124 126 127 133 138 139 141 143 145 91 97 99 101 103 106 109 110 117 118 119 120 122 122 RESULTS OF HEATING MILL NO. 1. One hundred per cent of all insects were killed on the second, third, and fourth floors. All insects were killed on the first floor, except those in elevator boots on the floor, where killing temperatures were not secured except directly over the radiators in the space beneath the floor. MiLi> No. 2. Brick building; day (.July 21, 1912) partly cloudy and calm; maximum out- door temperature 97° F. ; minimum outdoor temperature 74° F. Heating sys- tem, steam pipes along wall and a few radiators. Steam pressure of about 100 pounds maintained during heating. Capacity of mill, 1,000 barrels. BASEMENT AND FIRST FLOOR. Location and readvng of thermometers, — No. 1, in flour in an elevator boot resting on floor, 5 feet from steam pipes ; No. 2, hanging in the open, 6 feet high, 10 feet from steam pipes; No. 3, in flour in an elevator boot resting on floor, 15 feet from steam pipes ; No. 4, in flour in an elevator boot resting on floor, 10 feet from steam pipes ; No. 5, hanging in the open, 6 feet high, 15 feet from steam pipes; No. 6, in flour in a roll, 20 feet from steam pipes; No. 7, in flour in a roll in the cleaning room. 5 feet from a radiator. 34 BULLETIN 873, U. S. DEPARTMENT OF AGRICULTURE. Table III. — Data, showing rise in temperatures in basement and on first floor. Basement: Capacity, 33,790 cubic feet; radiation, 1,020 square feet First floor: Capacity, 40,040 cubic feet; radiation, 780 square feet. Time of day. Thermometer- No. 1. No. 2. No. 3. No. 4. No. 5. No. 6. No. 7. July 21, 1912: "F. 96 97 100 102 104 103 103 102 °F. 128 134 136 136 38 140 143 145 "F. 98 104 105- 107 108 109 109 113 °F. 98 98 98 102 106 109 109 109 °F. 130 138 140 140 142 143 °F. 121 123 123 124 125 127 °F. 97 10 a. m 100 12 m 102 2p. m 108 4 p. m 111 113 10.30 p. m 146 131 119 SECOND AND THIRD FLOORS. Location and reading of thermometers. — No. 1, hanging in ttie open, 6 feet high, 15 feet from .steam pipes; No. 2, in flour in bottom of an elevator boot, 15 feet from .steam pipes ; No. 3, in flour on bolting cloth in a reel, 12 feet from radiation (cleaning room) ; No. 4, hanging in the open, 5 feet high, 12 feet from steam pipes ; No. 5, in flour in a conveyor, 14 feet from steam pipes ; No. 6, hanging in the open, 5 feet high, 10 feet from radiator (cleaning room). Table IV. — Data showing rise in tempet'ature on second and third floors. Time of day. Second floor: Capacity, 43,120 cubic feet; radiation, 800 square feet. Third floor: Capacity, 43,120 cubic feet; radiation, 900 square feet. Thermometer- No. 2. No. 3. No. 4. No. 5. "F. °F. °F. 110 120 108 113 126 114 115 131 116 121 136 120 124 138 123 126 139 124 128 141 1'27 129 142 129 No. 6. July 21, 1912: 8 a. m 10 a. m... 12m 2 p. m 4 p. m 6 p. m 8 p. m 10.30 p. m 112 119 122 127 130 128 128 130 'F. 116 121 125 130 132 134 136 138 RESULTS OF HEATING MILL NO. 2. One hundred per cent of the insects were killed everywhere except in elevator boots resting on the concrete floor in the basement. Mill No. 3. Brick building; day (.July 27-28, 1912) calm and partly cloudy; outdoor maximum temperature 105" F. ; outdoor minimum temperature 73° F. ; heat- ing system, steam pipes along wall ; steam pressure of about 100 pounds main- tained during heating. Capacity of mill, 1,000 barrels. INSECT CONTROL IN FLOUR MILLS. 35 BASEMENT, FIRST, AND SECOND FLOORS. Location and reading of thennowrfers. — No. 1, hanging in the open, 5 feet high ; No. 2, in flour in an elevator boot resting on floor ; No. 3, in flour in an elevator boot resting on floor; No. 4, in some bran in a roll, 12 feet from steam pipes ; No. 5, hanging in the open, 5 feet high, 10 feet from steam pipes ; No. 6. in flour in an elevator boot resting on floor ; No. 7, hanging in the open, 5 feet high, 15 feet from steam pipes; No. 8, in flour in a conveyor, 20 feet from steam pipes ; No. 9, in 4 inches of flour near floor, 8 feet from steam pipes. Table V. — Data shomnff rises in temperatures in basement and on first and second floors. Basement: Capacity, 23,199 cubic feet: radia- tion, 260 square feet. First floor: Capacity, 29,526 cubic feet; radia- tion, 210 square feet. Second floor: Capacity, 27,417 cubic feet: radia- tion, 168 square feet. Time of day. Thermometer- No. 1. No. 2. No. 3. No. 4. No. 5. No. 6. No. 7, No. 8. No. 9. July 27-28: 9.30p. m 11.30 p. m July 28: (i.30a. m 7.30 a. m 9.30 a. m 11.30 a. m 12.30 p. m 2.30p.m 4.30p.m 5.30r).m 7.30 p. m 9.30 p. m 11 p. m "F. 113 112 109 110 118 122 122 124 128 129 130 130 130 134 136 136 'F. 104 102 96 99 108 109 110 113 114 115 119 117 118 120 122 122 °F. 102 102 100 99 110 109 110 113 115 115 118 117 118 ro 122 122 °F. 110 113 116 118 i:o 122 13 124 126 127 130 130 'F. 118 121 123 124 128 129 130 131 134 135 137 137 °F. 105 107 110 110 113 114 115 116 119 120 123 123 "F. 117 120 126 126 130 131 132 133 136 137 138 138 "F. 108 111 114 116 117 119 131 122 125 127 128 128 "F. 109 112 120 120 121 124 134 126 128 128 130 130 July 29: 1 a. m 3 a m 4 30 a m THIRD FLOOR AND DECK. Location and reading of thertnometers. — No. 1, in flour in a conveyor near the floor, 20 feet from steam pipes ; No. 2, hanging in the open, 5 feet high, 12 feet from steam pipes ; No. 3, in a pile of flour on the floor, 10 feet from steam pipes ; No. 4, hanging in the open, 5 feet high ; No. 5, in refuse flour on the base of a dust collector near the floor ; No. 6. in flour 3 feet from the floor. Table VI. — Data showing rise in tcmpcratnres on third floor and deck. Time of day. Third floor and deck: Capacity, 45,343 cubic feet, including deck; radiation, 326 square feet; radiation, deck, none. Thermometer- No. 2. July 27-28: 9.30 p. m. 11.30 p. m 6.30 a. m. 7.30 a.m. 9.30 a. m. 11.30 a. m 12.30 p. m 2.30 p.m. 4.30 p. m. 5.30 p. m. 7.30 p.m. 9.30 p. m. F. Ill 113 118 119 122 124 125 128 129 130 132 132 F. 12T 122 12S 130 132 136 137 139 141 142 )42 141 No. 3. F. Ill 113 119 120 122 124 126 128 129 130 132 131 3.4. No. 5. F. ° F. 123 113 124 115 130 120 131 122 134 126 K« 127 13S 128 141 133 143 133 114 134 144 136 143 136 No. 6. F. 115 116 122 124 126 129 1.30 134 136 137 138 1.38 36 BULLETTX 872, V. S. DEPARTMENT OF AGBTCULTITRE. RESULTS OF HEATI.NC MII.l. NO. .S. One hundred per rent 5 to 100 per cent were killed on the se<'ond, and DO to 95 per cent were killed on the first floor. 38 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE, Mill No. 5. Frame structure, not tightly constructed and with many small openings where lap-siding boards had sprung apart or warped. There are also small openings about doors. Building too loosely constructed for satisfactory fumi- gation with gas. Day (July 10-11, 1911) calm except for slight breeze blow- ing from southwest after 9 a. m. July 11, which somewhat affected temperatures on that side of building. Mill equipped with radiators for heating by steam during winter. Outdoor maximum tenipei'ature 95° F. No mention made of steam pressure or minimum outdoor temperature. FIRST AND SECOND FLOORS. Location and reading of thermometers. — No. 1, near airshaft ; No. 2, set 5 inches deep in flour ; No. 3, in boot of elevator ; No. 4, three feet from floor ; No. ."), on inside of outer wall; No. 6. near floor in elevator leg; No. 7, on support post ; No. 8, on outside wall ; No. 9, in a roll. Table XI. — Data showing rise in temperatures on first and second floors. First floor. Second floor. Time of day. Thermometer — No. 1. No. 2. No. 3. No. 4. No. 5. No. 1. No. 2. No. 3. No. 4. 7..30p. m ° F. 90 94 95 96 99 104 106 106 107 ° F. 86 91 95 103 104 107 109 110 111 " F. 100 102 106 108 112 113 114 115 116 " F. 102 124 127 131 134 141 141 141 ° F. 95 101 103 108 111 115 116 llfi ° F. 107 120 122 128 130 135 137 137 " F. 102 115 117 124 127 132 133 134 "> F. 101 109 116 118 120 125 126 124 ° F. 107 9. .30 p. m . 119 11.30 p. m 6 a. m 122 129 8 a. m. . . . 131 10 a. m 137 12 m . 137 2 p. m 138 3 p. m 141 IIS THIRD FLOOR AND DECK. Location and reading of tliermometers. — No. 1, in conveyor, in 3 inches of flour; No. 2, in elevator leg; No. 3, near outside wall; No. 4, in bolter; No. 5, on a post on the deck built on third floor. Tabe XII. — Data sJioiring rise in temperatures on third floor and deck. Time of day. 7.30 p.m. 9.30 p. m. 11.30 p. m 6 a. m 8 a. m 10 a. m. . . 12 m 2p. m No. 1. No. 2. No. 3. No. 4. No. F. 93 104 107 118 119 122 124 127 Thermometer— F. 104 113 118 125 129 134 135 136 102 113 116 125 129 138 136 136 F. 102 116 121 129 133 138 139 139 104 113 116 1'29 137 140 142 142 INSECT CONTROL IN FLOUR MILLS. 39 RESULTS OF HEATING MILL NO. 5. All insects were killed on the second and third floors. Dead beetles ajiid lurvse could be found beneath 3 inches of flour. All insects were killed on the first floor except near air shaft (thermometer No. 1) and those buried 5 inches in flour (thermometer No. 2). CONTROL BY FREEZING. If an infested mill is so equipped that it can be opened to low out- door temperatures without injury to equipment, freezin^: is an inex- pensive and valua.ble remedy for the flour moth. Turning off all heat and opening the mill to a temperature of zero or lower for from 3 to 10 nights continuously, or at intervals, has often proved effective in destroying the moth in its different stages. Mills in the northern United States and in Canada, where temperatures of 20 to 30 degrees below zero (Fahrenheit) are often experienced, have used this method of control with good results. Mills located farther south where freezing temperatures do not run so low or prevail continu- ously for any length of time have had little or no benefit from at- tempts to utilize cold weather. CONTROL BY SMUDGES. Control by means of smudges in the form of sulphur or tobacco fumes, etc., is depended upon rather extensively in many mills to reduce flour-moth injury. With smudges one can only expect to reduce the flour moth temporarily by killing the adult moths and certain of the immature stages. The fumes generated are not strong enough to kill as do hydrocyanic-acid gas fumigation and high tem- peratures. In rambling, loosely constructed buildings, where control by heat and hydrocyanic-acid gas is not practical, smudges, coupled with constant cleaning, have their value. The moth miller or adult is the most easily killed of the different stages of the flour moth and is the form killed in largest proportions by smudges. Since each female moth is the potential parent of several hundred larvae, one can appreciate the value of the persistent use of smudges. The chief drawback to dependence upon smudges is that they never completely exterminate the moth. Smudges, to accomplish satisfactory results,, must be applied frequently, and are therefore costly in the long run. One thorough application of heat rids a mill of the flour moth for long periods if provision is made against reinfestation from without. Flying moths may be found as soon as one day after the application of a smudge. CONCLUSION. Control of insect pests in flour and cereal mills has become a very important feature of food conservation and of mill construction and operation. Losses caused by mill insects to mill owners and the ad- 40 BULLETIN 872, U. S. DEPARTMENT OF AGRICULTURE. vantages of modern mill construction are working hand in hand to bring about a constant change for the better. Putting into practice methods of control advocated in this bulletin will make it possible for milling concerns to place upon the market a product reasonably free from infestation. Allien insect sanitation is conscientiously ap- plied, millers can state that blame often heaped upon them by brokers, retail grocers, and the public at large is due to one or more of several known conditions connected with warehousing and transportation, for the miller is in no way responsible. There are a number of serious insect pests of flour mills. The long- established pests in American mills do not interfere noticeably with operation of mill machinery. With the spread from Europe to the United States of the Mediterranean flour moth (p. 2-5). millers were forced to consider insect sanitation. The danger in the flour moth, sometimes called the " gray plague of flour mills,"" does not lie so much in food values actually consumed as in the enormous quanti- ties of silken threads spun by its larvae in and about mill machinery. This webbing habit causes the flour in passing through the machinery to form in ever-increasing clumps or webbed masses, which sooner or later choke or clog the machinery, and necessitates extensive mill shutdowns for cleaning and application of control measures. Experimental and practical demonstration work has proved the dependability of methods of control under certain conditions. These are fumigation with hydrocyanic-acid gas and the use of heat. Con- trol by freezing (p. 39) is less satisfactory. Smudges, as com- pared with fumigation with hydrocyanic-acid gas or the applica- tion of high temperatures, have only a temporary value. Preventive measures, including cleanliness, are of the greatest value in reducing losses due to insects. Dependence upon natural control by parasites is not advocated. The heat method is recognized as the most effective, practical, and inexpensive of all treatments and has the added ad- vantage of being absolutely safe. Where remedial measures nuist be applied in mills of moderate size, it has been estimated that the heat method is enough cheaper to pay in five years for the cost of the installation of enough radiation surface properly to heat the mill. Neither fumigation with hydrocyanic-acid gas nor the use of high temperatures, as recommended for mill-insect sanitation, injures the mill building or equipment or affects the baking qualities of flour. o !v. u^SKCfr UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 875 Contribution from the Bureau of Entomology L. O. HOWARD, Chief ^"W^^^?^^ Washington, D. G. July 20, 1920 COTTON BOLL WEEVIL CONTROL BY THE USE OF POISON.^ By B. R. CoAD, Entomological Assistant, and T. P. Cassidy, Cotton Entomolo- gist, Southern Field Crop Insect Investigations. CONTENTS. rage. Principles governing poisoning op- eration 1 Kind of poison to use 3 Poison specifications required — 3 Send samples of calcium arse- nate for analysis 4 Use of mixtures not recom- mended 4 Supply of calcium arsenate and dusting machinery available. 4 Keeping qualities of calcium ar- senate 5 Effect of the poison on man and animals 5 Plant injury by calcium arse- nate 6 How to apply poison 8 Amount required per acre for each application S Conditions under which to make applications 9 Arrangement of t>oisoning sched- ule 0 Season for applications 10 Time of starting poisoning 11 Time intei-val between applica- tions 12 Number of applications 14 Time to stop poisoning 14 Page. How to apply poison — Continued. Effect of rain on an application of poison 15 Starting poisoning in -the pres- ence of a complete infestation. 15 Earlier season treatment of isolated infestations 16 Organization of poisoning operation. • 17 Dusting machinery to use 18 Hand guns 19 Power dusters 21 WTieel-traction or cart dusters. 22 Need of an intermediate type of dusting machine 23 Lighting equipment for dusting machines 23 Capacity of machines for treat- ing several rows per trip 24 Features to be noted in purchasing cotton-dusting machinery 24 Hand gun 24 Wheel-traction or cart duster 25 Power duster 26 Cost of poisoning 26 Gains to be expected from poisoning.. 26 Advisability of ix)isoning under present conditions 28 Control of the cotton leafworm and fall army worm with calcium arsenate 29 PRINCIPLES GOVERNING POISONING OPERATION. IT SHOLTLD BE UNDERSTOOD that in poisoning for boll- weevil control extermination is not attempted or secured. The result aimed at is a sufRcie^it reduction of the weevil infestation to permit maturing a full crop of cotton. This is brought about by tak- ^ The investigations upon which this bulletin is based were in a sense the outgrowth of the work of Mr. Wilmon Newell, who published, together with Mr. G. D. Smith, in 174542°— Bui. 875—20 1 2 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. ing advantage of a combination of factors. In the first place, the cotton plant has a peculiar habit of putting on much more fruit than it is able to mature. About CO per cent of the squares which appear on the cotton plant fail to reach maturity as bolls, and are shed at some time during their development. This does not mean, however, that the cotton plant is shedding 60 per cent of its fruit throughout the season. This shedding is comparatively slight early in the season and increases rapidly as the plants develop until it reaches the point where all the new fruit which appears is shed. It has been found that up to a certain point the fruit shedding due to boll-weevil at- tack merely takes the place of this perfectly normal shedding which would be encountered even if the weevils were absent. The present system of weevil poisoning is intended merely to keep the weevils controlled to such a degree that they will not be able to do more than offset the normal shedding of the cotton plants. This means, generally speaking, that the weevils are permitted to multiply unmolested until they have become sufficiently abundant to puncture, more forms than would shed normally. Poisoning is then started and every effort is directed toward holding the infes- tation below this point of danger until the plants have had sufficient time to develop beyond weevil injury as many bolls as they will be able to mature. Then poisoning is stopped and the weevils are allowed to resume multiplication. Experience has shown that re- markably large increases in yield frequently result from a com- paratively slight degree of control for a short time during this critical period. It has also been very definitely determined, how- ever, that this effect is cumulative and can only be secured by start- ing the applications at the right time and repeating them at the correct time interval. This is the reason the writers urge " everyone contemplating weevil poisoning to decide upon conducting the op- eration as thoroughly as recommended or not to attempt it at all. Circular 33 of the Louisiana State Crop Pest rommisRion. an account of experiments in controlling the boll weevil by the application of powdered arsenate of lead which were conducted in Louisiana during 1908 and 1909. The studies of the U. S'. Department of Agricult\ire on the subject of cotton boll weevil poisoning were first described in Department Bulletin 731, July, 1918. The present bulletin gives the results of investigations carried on since that time. It is to be followed by a detailed bulletin giving descriptions of the various tests and studies which serve as a basis for the recommendations mude. The latter will also include a discussion of the contributions to the subject which have been made by various investigators for many years. The work has been conducted under the general direction of Dr. W. D. Hunter. The development of dusting machinery has been under the direction of Elmer .Johnson, of the Bureau of I'ublic Roads. The following have contributed to field and laboratory studies: F. F. Bondy, 11. W. Lee, T. F. McGeliec, R. W. Moreland, L. Z. Naylor, M. C. Rodgers, E. S. Tucker, W. B. Williams, and M. T. Young. The practical field tests have been rendered possible only by the hearty cooperation of the following : Prof. J. W. Fox, general manager, Delta & Pine Land Co. of Mississippi, Scott, Miss. ; Mr. Alex. Y. Scott, genernl manager, Charles Scott's Delta Plantations, Rosedale, Miss. ; and Mr. George S. Yerger, general manager, Maxwell-Yerger Planting Co., Mound, La. BOLL, WEEVIL CONTROL BY USE OF POISON. 6 KIND OF POISON TO USE. In the various studies which have been conducted by the writers, nearly every known type of arsenical has been tested, but the first generally successful results were secured from the use of an im- proved type of arsenate of lead. This was soon replaced by calcium arsenate, and this chemical remains the best which has been found for this purpose so far. Tests were conducted with both liquid and dry applications, and it was found that with the liquid spray a slight deg:ree of control was secured, but not nearly enough to make the operation profitable. POISON SPECIFICATIONS REQUIRED. When calcium arsenate was first tested it was found that the ma- terial as then prepared was not safe for use on cotton plants, owing to the injury caused by burning of the foliage. This was due to the excessive amount of water-soluble arsenic present. It was found that it was possible to make calcium arsenate without this high per- centage of soluble arsenic, and the type which is now recommended for this work is absolutely safe for use on plants. Not all calcium arsenate, however, is of this safe type. Anyone attempting this work should purchase the calcium arsenate on specifications describ- ing its composition. The specifications advised are as follows : Not less than 40 per cent arsenic pentoxid. Not more than 0.75 per cent water-soluble arsenic pentoxid. Density not less than 80 or more than 100 cubic inches per pound. Calcium arsenate was almost unknown as an insecticide when these experiments were inaugurated, and its i^roduction had been at- tempted by only a few manufacturers. Recently, however, almost all msecticide manufacturers have undertaken its production. While its manufacture is comparatively simple, some experience is required to produce a thoroughly satisfactory material. Undoubtedl}^ this difficulty will decrease rapidly as the manufacturers gain more ex- perience in the production of calcium arsenate and the quality of the material becomes standardized, but for the present it is still ad- visable to make purchases only on the specifications recommended. If the total arsenic content is too low, the material is not sufficiently poisonous to control the weevil. If the proportion of water-soluble arsenic is too high, plant injury will be the result. Some lots of calcium arsenate have been found which killed the cotton plants within a few hours after application. It is also important to watch the density of the material very closely. If it is too heavy and runs much less than 80 cubic inches to the pound, it is not suitable for use in dry powdered form and will not produce the proper type of dust 4 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTUEE. cloud to cover the cotton plants thoroughly. Likewise, if the mate- rial is too light and runs much over 100 cubic inches per pound, it is blown away from the plants too rapidly by any light air currents. SEND SAMPLES OF CALCIUM ARSENATE FOR ANALYSIS. The best way to make sure of having the proper grade of calcium arsenate is to send a sample to the Delta Laboratory at Tallulah, La., for free analysis. This analytical service was started by the United States Department of Agriculture last season and if possible will be continued until the qualit}^ of calcium arsenate on the market becomes more uniform. Every cotton planter who ])urchases calcium arsenate for this work is invited to send in samples and he will be given an immediate report as to whether or not the material is satisfactory for use in cotton dusting. For this purpose it is advisable to take at least two or three samples from as many different packages. Prob- ably the best plan is to sample 1 package in every 10 purchased. Half-pound samples should be taken from each and packed separ- ately. The complete record of each sample should be packed with that sample, giving its full history, including name of manufacturer, when purchased, size and tj^pe of package, condition of material, analysis claimed by manufacturer, etc. USE OF MIXTURES NOT RECOMMENDED. The only chemical now recommended for boll-weevil poisoning consists of calcium arsenate which conforms to the specifications de- scribed. Attempts have been made to utilize various mixtures of arsenicals or dilutions of calcium arsenate. It is quite possible that some of these may prove satisfactory in the future, but with our pres- ent information it is advisable to adhere strictly to the use of cal- cium arsenate without the addition of any diluent whatever. SUPPLY OF CALCIUM ARSENATE AND DUSTING MACmNERY AVAILABLE. Prior to 1919 the production of calcium arsenate was so limited that it was difficult to secure a sufficient amount to conduct the ex- periments desired in this work. During 1919, however, interest in its j)roduction vvas stimulated to the point where probabl}^ 3,000.000 pounds were sold for cotton- dusting work. It appears that a fairly ample supply of calcium arsenate will be available for 1920. This is especially true in view of the probable shortage of dusting ma- chinery. It has been found that successful results can not be se- cured unless special types of dusting machinery are utilized. These machines must be prepared for this particular purpose and their de- velopment has been considerably^ slower than that of the poison. At the present time the supply of machinery available is hardly suf- ficient to serve for the proper application of the poison already sold by the various manufacturers. BOLL WEEVIL CONTROL BY USE OF POISOIS". 0 KEEPING QUALITIES OF CALCIUM ARSENATE. Questions frequently are received as to whether or not calcium arsenate will deteriorate during long periods of storage. In some cases farmers purchased more than they could use during the past season and desired to hold over their surplus stock for use during the coming year. This is perfectly safe, as a properly made calcium arsenate should not deteriorate if stored in a reasonably dry place. Some lots have been kept at the Delta Laboratory for four years under normal storage conditions without the slightest deterioration. Moisture, however, even when it does not induce chemical deteriora- tion, will cause calcium arsenate to cake so badly that it can not be utilized in the dusting machines. EFFECT OF THE POISON ON MAN AND ANIMALS. Questions are frequently asked regarding the effect of calcium arsenate on laborers and mules engaged in the poisoning operations. It is not nearly as dangerous as Paris green, as it does not have the caustic action characteristic of the latter. There is, however, a cer- tain amount of danger attendant upon the use of any arsenical com- pound and reasonable precautions should be taken to protect the men and animals associated with it. Inhaling the dust should be avoided as far as possible. The most important precaution is that of personal cleanliness when using the material. The operators should be forced to bathe as soon as they complete the dusting work, and they should not be permitted to eat anything without at least wash- ing their hands and face thoroughl}^ No injury to work stock has ever been experienced during operations, but, it is safest to keep ordi- nary wire muzzles on all such animals in the poisoned fields. The question of fall grazing of poisoned cotton fields is frequently raised. As a rule the last application of poison is made at least a month, and usually two or three months, before the cotton crop is harvested and the animals allowed to graze on the cotton fields. Even if no rains have occurred to wash the poison from the plants during this period, the amount of leaf shed has been so great that the foliage of the plant has j^robably been renewed several times and the plants utilized for grazing purposes are largely, if not entirely, new growth. With regard to the danger to man involved in the use of calcium arsenate in the form of a dust cloud, it should be understood that arsenic may be taken into the body in three ways — by the mouth, by breathing, and by absorption through the skin. Ingestion or swallowing is the only manner which usually receives serious con- sideration, but in all probability it is the least important of the 6 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. three and can be avoided very easily by careful washing of the hands and face before eating or drinking. Inhalation is a danger which is constantly present and is difficult to avoid except by the use of dust masks or some type of filtering arrangement which removes the dust from the air entering the mouth and nostrils. Absorption undoubtedly is a very important means of taking arsenic into the human body. Even if extreme precautions are taken during dusting operations to avoid excessive direct contact with the poison supply, there is still a very considerable amount of poison dust settling all over the body from the air. This also has a decided tendency to adhere to the skin and " shed " water unless a strong lather is utilized, so a hasty rinsing in clear water should not be con- sidered as a satisfactory means of removing the poison. The danger of poisoning, although slight, should be considered. In the work conducted under the direction of the writers large quantities of calcium arsenate have been used by all sorts of laborers and generally with extreme carelessness. In spite of this, however, very few definite symptoms of even the slightest arsenical poisoning have been observed in connection with the field operations, and these few undoubtedly would have been avoided if proper precautions had been taken. Anyone using dry powdered calcium arsenate need not fear poisoning if he is reasonably cautious, but if any unexplained ill- ness should develop during its use the possibility of poisoning should be borne in mind and a physician consulted. The preliminary symp- toms of arsenical poisoning differ widely, but generally involve an intestinal and digestive disorder and are usually accompanied by some form of skin eruption. An additional indirect danger to live stock should be considered. The cloud of poison frequently has a tendency to drift considerable distances. Conditions may be such that a dangerous amount of poison will drift into a neighboring pasture and injure some of the stock feeding on the grass. Stock should certainl}^ not be allowed to graze on headlands and turnrows in fields which are being poi- soned. The same applies to chickens, turkeys, and other fowls. The only fatalities the writers have observed in the course of poisoning work have been in the case of chickens and turkeys running in poisoned fields and picking in the debris covered with a heavy dos- age of poison, where the machine had stopped and covered the ground rather thoroughly. PLANT INJURY BY CALCIUM ARSENATE. It is well to explain briefly the usual nature of arsenical injury to cotton plants, as there is a tendency on the part of inexperienced BOLL WEEVIL CONTROL. BY USE OF POISON. 7 operators to blame the poison for any form of plant disease. Plant injury by soluble arsenic ordinarily is termed " burning," and this word probably describes the appearance of the condition better than any other. The plants look very much as if they had been burned or scalded by some hot application, especially in cases of severe injury. The leaves and terminals droop and the young and more tender shoots wilt badly. This is followed very shortly by tlie death and drying of the most seriously injured tissue, causing lighter colored spots of dead tissue to appear over the leaves. In cases of mild injury this frequently involves only a sort of "shot-hole" ef- fect over the leaves, but when the injury is severe the entire leaf is killed and falls from the plant in a short time. In still more severe cases the entire plant may die soon after the application. One fact which should be borne in mind is that plant injury by soluble arsenic is generally very erratic and depends to a great ex- tent upon the weather conditions prevailing at the time of and directly following the application. It will be noted in the specifica- tions that a maximum limit of 0.75 per cent of w^ater-soluble arsenic oxid is recommended. This does not mean that all material con- taining more than this amount of water-soluble arsenic will be in- jurious to the plants at every application. In reality, applications of calcium arsenate containing as high as 3 or 4 per cent of water- soluble arsenic frequently will not burn the plants, but it has been found from a large number of tests that any material running over the maximum limit established in these specifications will burn the plants seriously ivJien certain iceather conditions are experienced. The most dangerous calcium arsenates are those averaging about 2 or 3 per cent in water-soluble arsenic, as it is possible to make two or three applications of these before any plant injury is experienced. The most dangerous conditions possible for plant burning are found during showery days when there are alternate showei-s and bright sunshine. On such, a day the plants are moist and the at- mospheric conditions ideal for making a satisfactory application of poison, but it has been found that if the water-soluble arsenic con- tent of the poison used is the least too high it will produce very severe injury. A nother point which should be noted is that very serious loss from plant burning may result from a degree of injury which appears very slight at first glance. In many cases the injui*y is such that only a portion of the leaf surface is burned, comparatively little leaf shedding is caused, and the plants appear to recover completely in a day or two. In reality, however, a very slight degree of leaf burning may assume serious proportions because it causes sufficient plant disturbance to produce a heavy shedding of fruit. 8 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. In case of leaf injury the cause of which is doubtful, it is ad- visable to pick several affected leaves, pack them in a moist wrapper, and send them either to the Delta Laboratory or to the plant pathologist of the nearest State experiment station. HOW TO APPLY POISON. AMOUNT REQUIRED PER ACRE FOR EACH APPLICATION. The quantity of calcium arsenate required for each application varies somewhat with the machinery utilized and with other condi- tions, such as the size of the plants, etc., but <»;enerally it has been found that at least 5 pounds per acre are necessary to make a satis- factory application. Where the machines are operated by experi- enced men, and especially by Government experts who have con- ducted this work for a number of years, it has been found that thor- oughly satisfactory results can be secured by an average application of about 4 pounds per acre. But where the work is conducted upon fi practical farm basis, the dosage used has averaged not less than 5 jjounds per acre. Anyone attempting the operation for the first time is more likely to average in the neighborhood of 7 pounds per acre for each application. As calcium arsenate at present prices aver- ages $0.25 per pound, any reduction in the amount required for each application is quite an important item, but to be on the safe side it is advised that until farmers are more thoroughly familiar with the operation the application should be excessive rather than too light. Properly conducted, the operation has a very large margin of profit and it is poor economy to risk this profit by attempting to save a pomid or two of poison per acre. The greatest factor for further saving in the amount of poison re- quired lies in the improvement of dusting machinery. The figures which have been quoted above apply only to the machines which are now in use in this work, and it is hoped that by further improve- ment of this machinery it will be possible to reduce the amount. In fact, considerable progT^ess has been made in this respect. When the first experiments were started with power dusting machines it was found impossible to control the weevils on an acre of cotton with less than from 12 to 15 pounds of poison dust. If the dust could be broken up into its finest particles and efficiently distributed, there would be enough material in about every 2 pounds of poison to dust an acre satisfactorily, but the present machinery is not equal to the task of delivering the dust in this form. This matter is being studied very carefully with the hope that further improved machinery can be devised. Ij BOLL WEEVIL CONTROL BY USE OF POISON. \) CONDITIONS UNDER WHICH TO MAKE APPLICATIONS. The time of day for makinoint, and thus exhaust the square supply and start attacking the young bolls. The majority of the poisoning operations in the past have been planned so as to start when about 15 to 20 per cent of the squares were punctured and then to repeat often enough to prevent the infestation from getting above about 25 per cent until the crop is set and the bolls are safe from weevil puncturing. In some cases where it is particularly desirable to confine the weevils to a certain cut and prevent any chance of migration, it is well to start at a somewhat lower percentage and continue even later, but where the only object in view is the benefit to the particular cut poisoned, there is apparently little to be gained from starting applications before at least 15 per cent of the squares are punctured. TIME INTERVAL BETWEEN APPLICATIONS. The question of the time interval between applications is very im- portant and one on which only conditional advice can be given, since it varies under different local conditions. In the past, once the ap- jolications were started they were generally made once a week as long as this seemed advisable or necessary. In reality, however, this selection of a time interval of one week was purely arbitrary and more recent results seem to indicate that in the majority of cases much better results can be secured by shortening this time interval. The effectiveness of a single application of calcium arsenate is de- cidedly limited in its duration. Its persistence on the plants natur- ally depends to a considerable extent on conditions prevailing at the time of application and immediately thereafter. In fact, while a very high percentage of control is secured during the first day of the application, this decreases the second day, and by the fourth day there is generally little or no effect. This short interval of effective- ness is due to two factors: (1) The poison is either washed or blown from the plants, and (2) new foliage is developed so rapidly that BOLL WEEVIL CONTROL BY USE OF POISON. 13 after a few days a great deal of unpoisoned tissue is present, which thus reduces the chances of poisoning the weevil. A four or five day time interval is best. This not only controls the brood of adults which were present svhen the applications were started, but also controls their progeny. Under average weather conditions the immature stages developing in the cotton forms at the time when an application is made will continue emerging over a period of at least 12 days. In other words, if the cotton can be kept poisoned thoroughly for a period of about 12 days, the weevils and their progeny are thoroughly controlled. Results vary of course with weather conditions but, generally speaking, two applications with a four-day interval have been found as effective as three or four applications with an interval of a week or more. If conditions are anywhere approaching normal about three applications with a four-day interval will usually reduce the weevil infestation to such an extent that it will be possible to stop poisoning, and in most cases this degree of control has been sufficient to persist during the re- mainder of the season. Another important point in this connection is that where a 7-day or 8-day interval is used and anything happens to interfere with the schedule, as it often will, the interval may be so extended that practically all control will be lost and the operation will become a complete failure ; whereas, when the shorter interval is attempted, any mishaps will still leave the applications sufficiently close together to give a fair degree of control and the worst that can be expected is the necessity of making one or two extra applications. It should also be taken into consideration that when the short in- terval is used the infestation can be permitted to become decidedly higher than would otherwise be safe. In other words, the control is so positive and the reduction of weevils so great that it is usually safe to permit the weevil practically to reach the point of becoming injurious to the cotton before making any application. The machinery requirement must be considered very carefully in connection with the time interval between ajjplications, as the acre- age which can be handled by any particular machine is proportion- ately decreased as the interval is shortened. This of course in- creases the investment in machinery required per acre. But this cost is offset by the decrease in the number of applications necessary and the consequent reduction in cost of poison and labor. It is offset to an even greater extent by the better control secured and the resultant higher yield of cotton. Of course at present the machinery supply is decidedly short and machines are now high in price compared to what they will be in the fviture when their production is standard- ized, so it is quite a problem to determine just how much the acreage allotment for each machine can be reduced without making the ma- 14 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. chine investment per acre too high. On the other hand, it is un- doubtedly the wisest poUcy to utilize methods which will give the greatest possible degree of control during the first few years when it is necessary for everyone to feel his way, more or less, in arranging the poisoning schedule. Consequently, it is urged that poisoning be attempted only under such conditions as will justify a sufficient machinery outlay to permit poisoning at about a four-day time interval. NUMBER OF APPLICATIONS. The question as to the number of applications is closely correlated with that of the time interval and is one which every man must de- cide for himself by experience, as the number will vary widely under different conditions. In the large Delta plantations, where the ma- jority of the work has been conducted in the past, it has been found that weevil control throughout the season in the cuts most heavily in- fested and requiring the earliest treatment in the spring has necessi- tated from four to six applications with a time interval of one week. Other cuts more distant from hibernation quarters and thus more lightly infested required three, two, one, or no applications. In two years' work en thousands of acres handled on this basis, the average number of applications for all cotton acreage involved has been something less than two. Where infestation is more uniformly dis- tributed at the outset and general poisoning is necessary, it has been found advisable to figure on an average of about four applications, and in case of an excessively rainy season, five or six applications. These figures are based on a one-week time interval, as this interval has been the one adopted in the majority of work so far. With a ; horter time interval, however, control can be secured with about three applications. This office is now planning to open a number of rmall experiment stations in nearly all representative districts of the cotton belt. By conducting control tests at all of these at the same time it w^ill soon be possible to designate much more definite rules of procedure under the varying conditions found in these different districts. TIME TO STOP POISONING. The time to stop poisoning has been more or less covered b}' the discussion of the number of applications but it is probably well to outline briefly the conditions governing this. The idea is to maintain weevil control below the point of loss to the crop long enough for the plants to set as inanj^ bolls as they can mature and also to protect these beyond the danger of weevil puncturing. Furthermore, there is nearly always a time when the plants cease to retain any more fruit. In many cases they not only discontinue squaring and blooming but shed the young bolls as fast as they are BOLL WEEVIL CONTROL BY USE OF POISON". 15 foitaed. This usually means that the plant has reached the limit of its ability to mature fruit. Naturally there is no advantage m attempting- to protect these forms, and poisoning during this period, if conducted after the retained l)olls are sufficiently large to escape weevil injury, is bound to be profitless. The time when this con- dition is reached varies widely with the season as well as with the soil fertility, and these factors must ahvays be taken into considera- tion in deciding whether additional poisoning is justified. EFFECT OF RAIN ON AN APPLICATION OF POISON. The effect of rain upon the application of calcium arsenate is a subject of much importance. Experience has shown that a certain amount of rainfall is desirable during the poisoning operations as it induces the formation of dew, makes conditions more nearly ideal for dusting, and apparently increases the amount of mortality se- cured from the applications. l*oisoning operations have been con- ducted during periods of extreme drought when there was almost no dew formation, but under such conditions it has been very difficult to secure a thorough degree of control. On the other hand, ex- cessive rain is detrimental, owing to the difficulty of getting the poison to stay on the plants long enough to control the weevils. This was especially true during some of the work in the extreme southern districts in 1919. The rains encountered in this work w^ere excessive and in many cases occurred almost daily. Under such conditions weevil poisoning can easily become an absolute im- possibility but, as far as that is concerned, such conditions make it practically impossible to raise cotton at all. It has been found that when an application is made under unfavor- able conditions and the plants are dry, even a slight shower occurring a short time later w411 wash off practically all the dust, while if con- ditions during the application are more favorable for a large quan- tity of the poison to adhere to the plants, a much heavier rain will not interfere seriously with the effect of the treatment. As a gen- eral rule it seems advisable to repeat an application immediately if a drenching rain falls within 24 hours after treatment. Until more information is secured on the subject, it undoubtedly will be best to follow some such rule as this, but the question of degree of weevil infestation and the conditions under which the application has been made should undoubtedly be considered in connection with the ques- tion of whether the application should be repeated. STARTING POISONING IN THE PRESENCE OF A COMPLETE INFESTATION. Many farmers make no move toward weevil poisoning imtil the crop is seriously infested and in fact almost totally destroyed by the 16 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. weevils. Then they want to get action immediately and desire to make a hurried purchase of calcium arsenate and dusting- machinery to try to save the situation. To secure control after the weevils have become excessively nu- merous and have punctured nearly all of the fruit is very difficult and usually involves the use of excessive dosages of poison at short intervals and, furthermore, requires an unusually large number of applications. Consequently the operation is very expensive at best and is also exceedingly hazardous. A very short delay under such conditions is extremely dangerous. The writers do not advise any- one to start poisoning under such conditions with a view of protect- ing the plants sufficiently to 'permit the setting of a neio crop. Sometimes a certain amount of poisoning under such conditions is extremely profitable, as, for instance, when a fair crop of young bolls has been set but is still in danger of weevil attack, owing to the fact that the bolls have not yet developed beyond the point of injury. When the weevils become excessively abundant in such a crop they quite frequently not only overtake the square formation but also destroy a high percentage of the bolls, and poisoning at this time will often save the bolls which are already present on the plants. Since these bolls require protection for only a very short period, it is possible to make perhaps two applications under such conditions and thus retard the weevils sufficiently to eliminate injury to the bolls. It should be remembered, however, that this is quite a dif- ferent proposition from starting in and attempting to control the weevils to the extent of permitting new squares to develop, bloom, form bolls, and reach maturity, EARLIER SEASON TREATMENT OF ISOLATED INFESTATIONS. The expense of weevil poisoning can be considerably reduced in many cases by localized treatment of the most heavily infested patches early in the spring. Often only a few acres in a large field will be infested early in the season. These may be distributed along a strip of timber or may adjoin barns, cabins, or other hibernation quarters. The weevils tend strongly to attack the largest plants available on emergence in the spring, and in case of any inequalities of soil fertility such as are very commonly found, it will generally be noted that the weevils will concentrate on the patches of larger cotton. Under any such condition it is often possible greatly to re- duce the infestation of the entire field and to delay considerably the date of general poisoning by going into these isolated spots fairly early in the season and treating them thoroughly before the weevils have an opportunity to spread. BOLL WEEVIL CONTROL BY USE OF POISON. 17 ORGANIZATION OF POISONING OPERATION. The organization of the poisonino- operation will differ consider- ably with variations in the size and type of the farm involved, but in general may be roughly classified into three groups : First, the large plantation operation where the poisoning is conducted by a separate organization on a wage basis, all crops being treated regardless of tenantry; second, the operation on a large plantation where the treatment of each crop is left to the individual tenant; third, the operation on the small fai-m w^here all the poisoning is conducted by the owner and his laborers. The majority of the w'ork wiiicli lias been conducted by this de- partment has been located on large plantations which are operated on a tenant basis. Under this form of operation each tenant is in a way an independent farmer in that he has a particular crop for which he is responsible, but at the same time he is operating under the direction of the plantation manager and deriving more or less support from the plantation, paying as a rule a certain proportion of the crop yield in lieu of rental. This condition complicates the poisoning situation. Some planters desire to purchase small-scale machinery for each tenant, leaving the treatment of his crop to the tenant and making this a part of the regular farming operations. Theoretically, the planter would seem by this arrangement to get the work done for nothing. In reality, however, the tenant's time is so fully occupied by his regular duties that he can conduct such work only by neglecting other necessary operations. Furthermore, it has generally proved impossible to get the operation properly conducted if left to the individual tenant. This not only means that the op- portunity for weevil control is lost on their crops but that these fields become a menace to the adjacent cotton which may have been properly poisoned. Weevil poisoning is a, flcmitatioiv mid not an in- dividual field' prop ositimi. For this reason nearly all the dusting- work which has been done in the past has been organized separately from the regular routine operations of the plantation. Separate la- bor is secured and assigned to the poisoning operations and applica- tions are made regardless of tenantry or crop arrangement, and are based purely on the distribution of the weevils over the place. This calls for skillful supervision, and where the property is large, involv- ing the use of a considerable organization, it is generally found advis- able to employ a competent man to take charge. This arrangement is particularly desirable at the present time, since the operation is so new and there is still so nmch to be learned about the most economi- cal means of procedure. The farmer who cultivates his own crop, or at least does so to a considerable extent, is in much closer touch with the progress of 18 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. events on his place and has a much smaller area to considei'. Under such conditions he can watch his cotton very thoroughly for weevil out- breaks and is usually fully informed on weevil distribution and abund- ance throughout his fields. Furthermore, under such conditions poi- soning is a comparatively minor operation, and by the use of hand guns or small-capacity traction dusting machinery it is possible for him to fit the weevil-poisoning v;ork into his regular operations and handle it with little or no interference. Of course, as a rule, his fields are smaller than on a large plantation and thus it is usually necessary for him to poison his entire crop. In a number of instances several neighbors have planned to enter into some agreement and purchase a traction-power machine coopera- tively with the idea of treating their several crops with the same ma- chine. Theoretically this arrangement seems a very good one, but there is considerable room for doubt concerning its practical work- ing. As has been pointed out, weevil poisoning is generally an emergency proposition and the work must be done at the right time or the entire operation may be imperiled. Consequently, the situa- tion which would develop in case the weather should interfere with a set schedule for machine operation on several places can easily be imagined. Crops belonging to the different men would require' treatment at the same time and it would simply be contrary to human nature if friction did not develop under such conditions. DUSTING MACHINERY TO USE. The selection of proper dusting machinery is undoubtedly equally as important as securing the correct poison. At the outset of the poisoning work an attempt was made to utilize or adapt existing types of dusting machines such as have been used for truck crops or orchard dusting, but it was soon found that cotton dusting required highly specialized machinery. All dusting operations which had been conducted previously had been in conjunction with more or less intensively handled crops where the labor supply per acre was generally rather plentiful and where the labor was of a fairly intelligent type. In the culture of cotton, however, dusting is placed on an extensive basis where it becomes a large field operation and the machinery must be made as efficient, fool-proof, and simple as possible. The success of cotton dusting is so dependent on thorough distribu- tion of the poison that any attempt to utilize a means of application which does not give this thorough distribution is certain to result in absolute failure. Many farmers have been accustomed to use the " bag-ancl-pole " method of poisoning for leaf worm control. Such a method of application is suited to leafworm control where large BOLL WEEVIL CONTROL BY USE OF POISON. 19 areas of plant tissue are devoured by the insects and the chances of poisoning considerably increased, but it will not furnish a suffi- ciently thorough distribution to result in control of the boll weevil. A special Farmers' Bulletin (1098) has been issued by the Depart- ment of Agi-iculture on the subject of dusting machinery for the cotton boll weevil, and everyone contemplating poisoning is advised to secure a copy of this and study it thoroughly. So far only three types of satisfactory dusting machines have been developed and placed on the market, namel}^, the hand gun, the wheel- traction machine, and the engine-power machine. HAND GUN. Several satisfactory models of hand gims are now on the market and may be purchased at from $15 to $25 each. Each machine con- sists of a small hand-operated fan and hopper slung from the shoul- der of the operator, which is carried through the field, poisoning a single I'ow of cotton at a time. Unfortunately these iiiachines are very difficult to operate. The labor involved is strenuous, to say the least, since the operator must walk through the field at a very fair pace and bear the strain of carrying, directing, and cranking the machine. It has been found that oj^erating for a short period of time, a man can cover about an acre an hour with one of these guns, but he can not continue at this rate for more than an hour or so. Owing to the necessity of poisoning when the plants are moist with dew, hand-gun work can usually be conducted only during the early morning or late evening. Generally speaking, this means from about 4 or 5 o'clock until 8 or 9 in the morning, and from about 6 o'clock to dark in the evening. About the best speed that can be attained is in the neighborhood of 5 acres per day for each man. If two men are available for each gun, it is possible to change occasionally and thus speed up the work so that the area covered by the single gun during the day is somewhat increased, but even under such conditions it is hardly safe to count on an average of more than 5 acres per day for each machine. Figuring on a four-day time interval, this would mean that each machine would cover about 20 acres, but in reality, with loss of time due to various causes which is certain to be ex- perienced, and with delay due to inability to operate under unfavor- able conditions, it is certainly not safe to figure on an average allot- ment of more than 15 acres to each hand gun. It would undoubtedly be much better to figure on about 10 acres for each hand gun, and, if possible, to distribute the operation of this gun between two dif- ferent laborers. In many cases farmers planting from 40 to 100 acres of cotton do not feel justified in purchasing large machines and desire to under- 20 BULLETIN 815, U. S. DEPARTMENT OF AGRICULTURE. take weevil poisoning by the use of a number of hand guns on this area. This has generally proved very unsatisfactory. The labor re- quirements of the guns are so great and the operation so laborious and difficult that labor troubles are certain to develop under any nor- mal condition and this has generally resulted in preventing anything like a satisfactory schedule for poisoning work. One important use of the hand gun is in conjunction with power machinery. A great many fields have certain portions which are very difficult to treat with power machinery, owing to the presence of ditches, stumps, short rows, or similar obstacles. Attempts to treat these portions of the fields with large machinery greatly reduce the rapidity of the operation of the machine and frequently the cot- ton is injured by being tramped and driven over. Under such condi- tions the efficiency of the large machine is much increased if a few hand guns with which to treat these difficult portions are available for use at intervals. Hand guns are also of great value for the early-season work in treating isolated spots of infestation which are often comparatively small in extent. In much of the work which has been conducted in the past, hand guns have been used in such fields for the first appli- cation or two and a sufficient degree of control secured from this work to make it possible to defer starting general poisoning with a large machine until some weeks later than would otherwise have been the case. Several instances have been noted where hand machines have been offered for cotton dusting purposes with the distributing system di- vided between two nozzles, the idea being that two rows could be treated at each trip. None of the machines manufactured at present have more than enough fan power to treat a single row satisfactorily. A number of hand guns have been offered for use in cotton dusting with the blower constructed so that the dust delivery is intermittent. As such guns are usually equipped with a bellows serving as a blower the powder is expelled only in recurrent blasts. These are not satis- factory for cotton work since it is necessary that the gun discharge the dust continuously in order that uniform poisoning of all plants be secured while the operator proceeds along the row. As has been stated, hand-gun work is mainly conducted in the early morning and late evening, and this limits very much the t^mount of territory which can be covered each day. Operation dur- ing the night, however, is often entirely feasible and the results from such operations are usually better than where treatments are attempted during the day. Very satisfactory hand dusting can be done at night by the aid of a small oil or carbide light attached to the hat of each operator, or, in some cases, several hand guns may be BOLL WEEVIL CONTROL BY USE OF POISON. 21 worked together as a group if some kind of torch or contractor's flare is placed at the row ends to provide light. Another way of reducing the amount of labor involved in hand dusting is hy the operation of the machine from muleback. This has been tried on numerous occasions and is generally much more satisfactory than by walking. POWER DUSTERS. The earliest work on weevil poisoning was conducted entirely with hand guns, as very small areas were treated during the strictly plat-test stage of the work. Ordinarily the next step in develop- ment of dusting machinery would have been to produce something of slightly larger capacity, but owing to the rapid development of the work at this stage, attention was transferred from the develop- ment of hand gims to that of blowers of the largest possible capacity, namely, the engine-power machines. In this case the duster was a horse-drawn machine with the fan and feeder operated by a small gasoline engine mounted on the platform. This machine had a dis- tributor extending across the rear end with five nozzles spaced 4r| feet apart, thus covering approximately five rows. Several models of these machines were devised and placed on the market and were used rather extensively during 1918 and 1919. It was found that they would cover from 6 to 10 acres an hour while in operation but that the loss of time clue to mechanical difficulties was so gi'eat that a machine seldom averaged over 40 or 50 acres for the day's opera- tions. Continued use of these machines soon made it obvious that they were too complicated and cumbersome to be thoroughly satis- factory for cotton-dusting work. Probably the most serious diffi- culty was the gasoline engine. Another great difficulty was found in constructing a thoroughly satisfactory distributing system for spanning five rows. There was considerable length of pipe extend- ing out beyond the machine on both sides and as this was necessarily made very heavy, the weight caused almost constant breakage due to excessive vibrations and jarring. In many cases the distributors of these power dusters were cut off so as to cover only three rows at a time and they really treated more acreage throughout the season than when they were arranged to span five rows, owing to the greater convenience in handling the machine over the field, around stumps, fences, etc., and the decreased loss of time from breakage. These power machines were placed on the market at prices ranging from $400 to $600 and this price seemed high for the work they could accomplish. Consequently it was desirable to devise a machine which would eliminate the gasoline engine and span only three rows 22 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. with its distributor and which could be sokl at a more reasonable price. The only answer to this problem which has been developed so far is the wheel-traction or cart type of duster. WHEEL-TRACTION OR CART DUSTERS. For the wheel-traction duster a light two- wheel cart is utilized and the power for driving the fan, feeder, etc., is derived from the wheels of the cart. This duster is pulled by two mules and is operated b}^ one man instead of two. By the beginning of 1919 a tentative model had been produced which was placed in the field and operated on a practical basis throughout that season. This machine proved emi- nently satisfactory in spite of the fact that it was very crudely con- structed, and it was selected as the most desirable type for general use in cotton dusting on a large scale. As built, it proved far simpler and easier to operate than the power machines and was especially A^aluable for its convenience in driving through the field. A number of duster manufacturers became interested in this type of machine and several models based somewhat on this idea are now on the market and others are in the course of construction. These differ widel}^ and embrace the ideas of individual designers, but the general principles of construction are more or less the same. Practically all of the machines are built with an arched axle providing a 42-inch clearance beneath the arch and having a tread of about 48 inches, thus enabling satisfactory operation in any cotton rows not narrower than about 36 inches. It has been found that under normal conditions one of these wheel-traction machines will probably average about 25 acres per night of operation. It may exceed this amount under very favorable conditions but it is not safe to figure on a greater average than this. As has been pointed out, it is undoubtedly desirable to provide equipment for treatment at a time interval of about 4 days. Owing to rains and other interruptions, it is undoubtedly not safe to figure on more than three days' operation out of the four. Consequently, about the best that can be expected of one of these machines is that it will handle in the neighborhood of 75 acres of infested cotton throughout the season. So far the machines of this type which have been built have "been placed on the market at a rather high price, ranging from about $350 to $500. The manufacturers have taken extreme precautions to place the highest quality of workmanship and material in the con- struction and this has resulted in making the price of the machine higher than is generally desirable. It still remains to be seen whether or not this cost is justified by increased efficiency of the machines. It seems quite probable that other machines built on BOLL WEEVIL CONTROL BY USE OF POISON. 23 much the same design will be placed on the market at a fairly early date at a lower figure. From present prospects the minimum sales price for such a machine will be in the neighborhood of $200. The high cost of this machinery is particularly unfortunate, as it will tempt the farmer to expand his acreage allotment for each ma- chine as much as possible by increasing the time interval between ap- plications. This is bound to be a hazardous practice. In fact, if a grower is confronted with the problem, it would be far better for him to select a limited acreage of his best yielding land where the weevil damage is highest and treat this thoroughly and properly, ignoring the rest of his crop, than risk the success of the entire operation by attempting to make a machine cover too much acreage. NEED OF AN INTERMEDIATE TYPE OF DUSTING MACHINE. As the situation now stands, there is no dusting machine on the market intermediate between the hand duster and the wheel-traction or cart type of duster. Inasmuch as hand dusters are generally un- satisfactory on fields larger than about 25 acres, and as the cost of a cart duster is such that a man is seldom justified in buying one for use on less than 75 acres, no equipment suitable for a man cultivating between 25 and 75 acres of cotton is available and his problem is a difficult one. When a sufficient supply of labor is available, however, it may be possible for him to utilize hand guns; and where the soil fertility is particularly high, the purchase of a cart machine for use on the small acreage may be justified. Several types of devices, such as saddle guns or single-wheel machines for operation between the cotton rows, are being studied, but the process of perfecting and producing them will require considerable time. LIGHTING EQUIPMENT FOR DUSTING MACHINES. The question of lighting any cotton-dusting machine requires par- ticular attention. As has been explained, an application of poison gives the best results when it is made at night ; therefore a thorough lighting equipment must be provided. This has been a serious diffi- culty in the past and all kinds of lights have been tested. So far only one type has been developed which gives any assurance of being thoroughly satisfactory. This is a special model of acetylene light which utilizes a compressed carbide cake for fuel. These lights have been constructed for cotton-dusting machine work and apparently the majority of the machines to be placed on the market will be equipped with them. At any rate, anyone purchasing a dusting machine should be sure that it is provided with a simple and satis- factory equipment that will give ample light for avoiding stumps, 24 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. ditches, etc., while driving. It must also afford sufficient rear illu- mination over the nozzles to enable the operator to make sure that the dust flow is regular. CAPACITY OF MACHINES FOR TREATING SEVERAL ROWS PER TRIP. In the earlier stages of the work it was attempted to make machines with adjustable nozzles that could be aimed directly at the rows being treated. It was soon found, however, that cotton rows vary so greatly in spacing that such an arrangement was impractical. Moreover, it was unnecessary. The entire poisoning operation is based on the plan of creating a thick cloud of dust which will en- velop all parts of the cotton plants and coat them thoroughly. Con- sequently, all width-adjusting devices were eliminated and the noz- zles have since been spaced about 4^ feet apart. These nozzles are provided with deflecting plates or some similar spreading device which causes the dust clouds from the different nozzles to unite very quickly after leaving the machines, so that a uniform " fog " ex- tending from row to row is created. Under such conditions it does not matter whether the nozzles are directly over the rows or between them. It has also been found possible under certain conditions to cover more rows than the nozzles actually span. This suggestion, however, should be taken very conservatively, as considerable ex- perience is required in the use of dusting machinery to determine just how far the drift can be considered as thoroughly effective in dusting. Until the operators have become thoroughly experienced in this work it is probably the best plan to figure on taking only as many rows as the distributor will actually span. FEATURES TO BE NOTED IN PURCHASING COTTON-DUSTING MACHINERY. For the benefit of those planning to purchase cotton-dusting ma- chinery, the following brief outlines have been prepared giving the most important features which should be considered in order to make sure that the machines will be satisfactory'. HAND GUN. The total weight should be not over 20 pounds when filled with poison dust. The hopper should hold about 4 to 7 x^ounds of calcium arsenate and it should be possible to put out practically all of this before refilling. The balance of the gun should he such as to cause the least strain on the oj)erator, that is, the heavy parts of the machine should be as close to his body as possible. BOLL WEEVIL CONTROL BY USE OF POISON. 25 The gun should haneTatlon involved in the production of the cotton crop just hecause the plants are hehig poisoned. Weevil poisoning can not make cotton. It is up to the farmer and the land to do this, and the best that can be expected of poisoning is to save this cotton from weevil destruction. CONTROL OF THE COTTON LEAFWORM AND FALL ARMY WORM WITH CALCIUM ARSENATE. One question which frequently arises in connection with the use of calcium arsenate is whether or not this material will control the cotton leafworm, fall army worm, or any other pests of this nature. It will undoubtedly be as satisfactory for this purpose as any chemical which could be utilized. It is very nearly as poisonous as Paris green to the worms and has the decided advantage of being cheaper and less injurious to the plants. In case anvone desires to utilize weevil-poisoning equipment solely for leafworm control, however, he should bear in mind that he could considerably reduce the expense of the operation without interfering with its effective- ness by mixing equal parts of lime and calcium arsenate and apply- ing this mixture at the rate of about 4 or 5 pounds per acre. 30 BULLETIN 875, XT. S. DEPARTMENT OF AGRICUULTURE. PUBLICATIONS OF UNITED STATES DEPARTMENT OF AGRICUL- TURE RELATING TO INSECTS AFFECTING COTTON PLANT. AVAILABLE FOR FREE DISTRIBUTION. Cotton Improvement under Weevil Conditions. (Farmers' Bulletin 501.) Fall Army Worm, or " Grass Worm," and Its Control. (Farmers' Bulletin 752.) Carbon Diwulphid as an Insecticide. (Farmers' Bulletin 709.) Red Spider on Cotton and How to Control It. (Farmers' Bulletin 831.) Bollworm or Corn Earworm. (Farmers' Bulletin 872.) How Insects Affect the Cotton Plant, and Means of Combating Them.' (Farm- ers' Bulletin 890.) Relation of the Arizona Wild Cotton Weevil to Cotton Planting in tlie Arid West. (Department Bulletin 233.) Studies of the Mexican Cotton Boll Weevil in the Mississippi Valley. (Depart- ment Bulletin 358.) Argentine Ant: Distribution and Control in the United States. (Department Bulletin 377.) Cotton Boll Weevil Control in the Mississippi Delta, with Special Reference to Square Picking and Weevil Picking. (Department Bulletin 382.) Collection of Weevils and Infested Squares as a Means of Control of the Cotton Boll Weevil in the Mississippi Delta. (Department Bulletin 564.) Pink Bollworm with Special Reference to Steps Taken by the Department of Agriculture to Prevent Its Establishment in the United States. (Depart- ment Bulletin 723.) Recent Experiments in Poisoning Cotton Boll Weevils. (Department Bulletin 731. ) FOR SALE BY THE SUPERINTENDENT OF DOCUMENTS, WASHINGTON, D. C. Use of Paris Green in Controlling the Cotton Boll Weevil. 1904. (Farmers' Bulletin 211.) Price 5 cents. Controlling the Boll Weevil in Cotton Seed and at Ginneries. 1904. (Farmers' Bulletin 209.) Price 5 cents. Miscellaneous Cotton Insects in Texas. 1905. (Farmers' Bulletin 223.) Price 5 cents. Control of the Cotton Boll Weevil. 1912. (Farmers' Bulletin 500.) Price 5 cents. Boll Weevil Problem with Special Reference to IMeans of Reducing Damage. 1917. (Farmers' Bulletin 848.) Price 5 cents. Recent Studies of the Mexican Cotton Boll Weevil. 1914. (Department Bul- letin 231.) Price 5 cents. Studies on the Biology of the Arizona Wild Cotton Weevil. 1916. (Depart- ment Bulletin 344.) Price 5 cents. Red Spider on Cotton. 1917. (Department Bulletin 416.) Price 20 cents. Mexican Cotton Boll Weevil. 1897. (P^ntomology Circular 18.) German and Spanish editions, each 5 cents. Most Important Step in Control of Boll Weevil. 190S. (Entomology Circular 95.) English and French editions, each 5 cents. What Can be Done in Destroying Cotton Boll Weevil During Winter. 1909. (Entomology Circular 107.) Price 5 cents. Status of Cotton Boll Weevil in 1909. 1910. (Entomology Circular 122.) Price 5 cents. BOLL WEEVIL CONTROL BY USE OF POISON. 31 Annotated Bibliograpliy of Mexican Cotton Boll Weevil. 191\ (Entomology Circular 140.) Price 5 cents. Movement oi' Mexican Cotton Boll Weevil in 1911. 1912. (Entomology Circu- lar 146.) Price 5 cents. Two Destructive Texas Ants. 1912. (Entomology Circular 148.) Price 5 cents. Cotton Stainer. 1912. (Entomology Circular 149.) Price 5 cents. Cotton Worm or Cotton Caterpillar. 1912. (Entomology Circular 153.) Price 5 cents. Movement of Cotton Boll Weevil in 1912. 1913. (Entomology Circular 167.) Price 5 cents. Report on Habits of Kelep, or Guatemalan Cotton Boll Weevil Ant. 1904. (Entomology Bulletin 49.) Price 5 cents. Mexican Cotton Boll AVeevil, Revision and Amplification of Bulletin 45, to Include Important Observations made in 1904. 1905. (Entomology Bul- letin 51.) Price 15 cents. Report on Miscellaneous Cotton Insects in Texas. 1906. (Entomology Bul- letin 57.) Price 5 cents. Proliferation as Factor in Natural Control of Mexican Cotton Boll Weevil. 1906. (Entomology Bulletin 59.) Price 15 cents. Papers on Cotton Boll Weevil and Related and Associated Insects. 1909. (Entomology Bulletin 63, 7 pts.) Price 15 cents. Hibernation and Development of Cotton Boll Weevil. 1907. (Entomology Bulletin 63, pt. I.) Price 5 cents. Notes on Biology of Certain Weevils Related to Cotton Boll Weevil. 1907. (Entomology Bulletin 63, pt. IJ.) Price 5 cents. Ant Enemy of Cotton Boll Weevil. 1907. (Entomology Bulletin 63, pt. III.) Price 5 cents. Predatory Bug Reported as Enemy of Cotton Boll Weevil. 1907. (Entomology Bulletin 63, pt. IV.) Price 5 cents. Cotton Stalk-borer. 1907. (Entomology Bulletin 63, pt. VII.) Price 5 cents. Mexican Conchuela in Western Texas in 1905. 1907. (Entomology Bulletin 64, pt. I.) Price 5 cents. Notes on Economic Importance of Sowbugs. 1907. (Entomology Bulletin 64, pt. II.) Price 5 cents. Studies of Parasites of Cottim Boll Weevil. 1908. (Entomology Bulletin 73.) Price 10 cents. Some Factors in Natui'al Control of Mexican Cotton Boll Weevil. 1907. (Ento- mology Bulletin 74.) Price 15 cents. Hibernation of Mexican Cotton Boll Weevil. 1909. (Entomology Bulletin 77.) Price 25 cents. Plant-bugs Injurious to Cotton Bolls. 1910. (Entomology Bulletin 86.) Price 20 cents. Insect Enemies of Cotton Boll Weevil. 1912. (Entomology Bulletin 100.) Price 15 cents. Argentine Ant. 1913. (Entomology Bulletin 122.) Price 25 cents. WASHINITOX : GOVERNMENT PRINTING OFFICE : 1920 UNITED STATES DEPARTMENT OF AGRICULTURE BULLETIN No. 885 Contribntion from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. December 11, 1920 THE BLACK FLY OF CITRUS AND OTHER SUBTROPICAL PLANTS By HARRY F. DIETZ, Entomological Inspector, Bureau of Entomology, United States Department of Agriculture, and JAMES ZETfEK, Entomologist, the Panama Canal CONTENTS Introductory 1 The Scientific Name of the Insect and to Whom it should be Accredited ... 2 Literature 3 Origin and Distribution 3 Spread of the Insect in the New World . 6 Food Plants 14 Injury 18 Life History and Habits 22 Eggs 29 First Larval Instar 31 Second Larval Instar 32 Third Larval Instar . 33 Fourth Instar, or Pupa 34 Page Life History and Habits— Continued. Fifth Instar, Adult or Imago ... 36 Technical Description , . 39 Seasonal History 42 Parthenogenesis 44 Natural Factors that Tend to Control the Black Fly in the Canal Zone .... 44 Natural Enemies 45 Artificial Control 45 Possibility of the Black Fly being Intro- duced into the United States and Fac- tors Influencing its Establishment Here 47 Summary 52 Literature Cited 53 WASHINGTON GOVERNMENT PRINTING OFFICB 1920 UNITED STATES DEPARTMENT OF AGRICULTURE , BULLETIN No. 885 Contribution from the Bureau of Entomology L. O. HOWARD, Chief Washington, D. C. December 11, 1920 THE BLACK FLY OF CITRUS AND OTHER SUB- TROPICAL PLANTS.^ By Harry F. Dietz, Entomological Inspector, Bureau of Entomology, United States Department of Agriculture, and James Zetek, Entomologist, The Panama Canal.^ CONTENTS. Introductory 1 The scientific name of the insect and to whom it should be accredited 2 Literature 3 Origin and distribution 3 Spread of the insect in the New World 6 Food plants 14 Injury 18 Life history and habits 22 Eggs 29 First lar\-al instar 31 Second larval instar 32 Third lar\'al instar 33 Fourth instar, or pupa 34 Life history and habits— Continued. Fifthinstar, adult or imago 36 Technical description 39 Seasonal history 42 Parthenogenesis 44 Natural factors that tend to control the black fly in the Canal Zone 44 Natural enemies 45 Artificial control 45 Possibility of the black fly being introduced into the United States and factors influ- encing its establishment here 47 Summary 52 Literature cited 53 INTRODUCTORY. The black fly, Aleurocanihus woglumi Ashby, known also as the spiny citrus white fly and as the mosca prieta, is a tropical pest of the > This work was done in cooneration with the Panama Canal, the second-named author being delegated to work with the bureau's representative. To the various departments and divisions of the Panama Canal which have aided and cooperated with them the authors are grateful. Special thanks are due the health de- partment for providinglaboratory space and facilities for carrying on the work, and the authors are indebted to the cliief health ofllcers. Col. H. C. Fisher and Lieut. Col. A. T. McCormack, and to the directors of the Board of Health laboratories, Drs. Oscar Teague, Wm. L. McFarland, and Lewis B. Bates, for their en- couragement and advice from time to time. The studies on the life history, habits, hosts, distribution, and control of the black fly in the Canal Zone and the adjoining parts of the Republic of Panama were begun in Jime, 1918, and tlie authors have been assisted in this work by Mr. Ignacio Molino, who, as entomological laboratory assistant, deserves special mention for the way he has carried on the work that has been assigned to him. Observations and experi- ments made by him are given full credit in the following pages. Practically all the pubUshed literature relating to Aleurocanihus woglumi has been consulted and the data set forth compared with those obtained in the Canal Zone. Unpublished reports on this insect in the files of the Bureau of Entomology have been consulted, and much that was both interesting and helpful has been obtained in the discussions that one or both of the authors have had with Mr. J. R. Johnston, of theComision de Sanidad Vegetal de Cuba, and Mr. Harold Morrison and Dr. A. C. Baker, of the Bureau of Entomology. Full use has been made of an unpublished report of Mr. Morrison on the status of the black fly in the West Indies. » The arrangement of the authors' names is alphabetical and denotes neither seniority nor precedence. 185528°— 20 1 2 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. Old World that has made its appearance in the Tropics of the Amer- icas. Its apparently rapid spread in the New World along with the fact that it now threatens to gain entrance into Florida from Cuba and Nassau makes it a pest of special interest from the plant quar- antine point of view. In August, 1917, Mr. Harold Morrison, the explorer in charge of the Mediterranean fruit fly investigations of the Bureau of Ento- mology, found the insect well established in Cristobal, Balboa, and Ancon in the Ca^ial Zone, and in Colon and Panama in the Republic of Panama. Since the Canal Zone may be regarded as the crossroads of the New World's water traffic, a temporary field station was estab- lished at the Board of Health laboratory at Ancon to study the black fly and other tropical insect pests that might be carried by commerce to other parts of the Tropics or to the subtropical parts of the United States. Aleurocanthus woglumi has several common names. It has been called the black fly, the citrus black fly, and the black scale in Ja- maica and is known as the blue fly and the citrus blue fly in the Bahamas. In Florida it is known as the black fly. In literature it has also been referred to as the spiny citrus white fly in order to distinguish it from the common citrus white fly {Dialeurodes citri Ashmead). While this insect belongs to the family Aleurodidae, known as white flies, the term of white fly does not fit it, inasmuch as it is black or dusky in all stages. In Spanish countries the insect is known as mosca prieta, which means black fly. There is a minor objection to the use of the name black fly, inasmuch as it is the common designation of biting buffalo gnats, family Simuliidae. Nevertheless, on account of the common and wide use of the name "black fly" for this aleurodid in English-speaking countries and of its Spanish equivalent in Spanish-speaking countries, it would probably be impossible now to secure the general adoption of any other name. There is little likelihood of any confusion arising in discussion or literature in the use of the term black fly for this insect as an enemy of citrus and other subtropical plants. THE SCIENTIFIC NAME OF THE INSECT AND TO WHOM IT SHOULD BE ACCREDITED. The black fly was first described by Ashby (3)' in his article on the black scale or black fly. Although this description is entirely untechnical and is used merely in setting forth the stages in the life cycle, and although Ashby used Quaintance and Baker's manuscript name for the insect, and specifically states that this name should be credited to Quaintance, nevertheless, under the rules of the Inter- * Figures In parentheses refer to " Literature cited," p. 53-55. THE BLACK FLY OF CITRUS. 3 national Code of Nomenclature, Ashby's name must stand as the author of the species. Quaintance and Baker (31) have written the first and only technical description of Aleurocanthus woglumi, although the material that they had at hand did not permit their describing the first to third larval instars of this insect. The com- plete descriptions of all stages of the insect prepared by one of these authorities are included in this report. LITERATURE. There are 33 references to Aleurocanthus woglumi in literature. Among the most important of these is the work of Johnston (16, 17), Cardin (7-11), and Hutson (12-15) in Cuba; Ashby (2,3) and Ritchie (33, 34) in Jamaica; and that of Quaintance and Baker (31, 32) in the United States. Others who have contributed to the literature are Arango (1) in Cuba; Ballou (4) in Barbados; Bragdon (5), Montgomery (19), Newell (22-29), Pierce (30), and Watson (35) in the United States; McCormack (18, p. 30), in Panama; and Zetek (37) in Costa Rica. An unpublished work by Morrison in the files of the Bureau of Entomology coverte his observations in Jamaica and Cuba and records his finding the insect for the first time in the Canal Zone. This report has been freely consulted and used by the authors. No comprehensive publication on the black 11}', however, has appeared to date. A bibliography of the literature cited is given on pages 53 to 55. ORIGIN AND DISTRIBUTION. Aleurocanthus woglumi unquestionably originated in the East Indies and from there it has been introduced into the Tropics of the New World, over which it is now spreading. It was first found in India by Maxwell-Lefroy in June, 1910. In the same year it was found in Manila, P. I., by George Compere. In October, 1910, R. S. Woglum, of the Bureau of Entomology, in search of parasites of the citrus white fly (Dialeurodes citri Ashmead), found Aleurocanthus woglumi at the Royal Botanical Gardens, Ceylon, and it is in his honor that the species is named. From November, 1910, to June, 1911, Woglum found this insect at the following places in India: Gujranwala; Kalim- pong, Sikkim; Lahore and Nagpur, Central Province, In November, 1913, Mr. A. Rutherford added Peradeniya, Ceylon, to the Old World distribution of the black fly. Aleurocanthus woglumi was first sent in to the Bureau of Entomology for determination from the New World in November, 1913, and Feb- ruary, 1914, by Col. C. Kitchner, from Half -Way, Jamaica. In May, 1914, S. F. Ashby, microbiologist of the Department of Agricul- ture of Jamaica, sent in specimens of tliis pest from Kingston, and BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. in 1915 he wrote two papers (2) and (3), indicating therein that it was rather widespread over the island of Jamaica. In February, 1916, Patricio Cardin, entomologist of the experiment station of Cuba, sent in specimens from Guan- tanamo, Cuba, and in the same month L. J. K. Brace sent in specimens from Nas- sau, New Providence, Ba- hama Islands. In 1 9 1 7 John- ston (16, 17) showed that the insect occurred in Cuba, at Guantanamo and Habana. This same year Ritchie (34) states that ' ' the entire island [Jamaica] is becoming gener- ally involved." In an un- pubhshed report of Novem- ber, 1917, in the bureau files, Harold Morrison gives an ac- count of the black fly in Jamaica and Cuba, and for the first time called attention to the fact that tliis insect was established in the Canal Zone and the adjoining parts of the Republic of Panama, having been found by him on citrus and mango trees in Cristobal, Ancon, and Balboa in the former place and in Colon and Panama in the lat- ter, in August of that year. In the early part of 1918 Dr. W. M. Mann, of the Bureau of Entomology, foimd the pest in the vicinity of Santi- ago de Cuba, thereby estab- lishing a new Cuban distribu- tion record. In his explora- tions of the British West THE BLACK FLY OF CITRUS. Indies in 1918, Morrison did not find this pest in Barbados, Grenada, Tobago, Trinidad, or 'British Guiana. The British entomologists of these regions are fully acquainted with this insect and are carefully Fig. 2.— Distribution of the black fly in the West Indies. watching for it. Therefore it is safe to assume that the black fly does not occur there at the present time. In his trip to the Virgin Islands in 1917, Morrison did not find it at any places visited by him ,S?<^^y^9Gey • Fig. 3. — Distribution of the black fly in Cuba. on the islands of St. Croix, St. John, or St. Thomas, Neither did he find it in Porto Rico in his inspections that year. In 1919, Zetek (37) records for the first time the occurrence of the black fly in Costa Rica, having found it in the vicinity of Limon and as far inland as Peralta, on the Northern Railroad. The maps show the distribution of the black fly in the world (fig. 1), in the West Indies (fig. 2), and in Cuba (fig. 3). 6 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. In the Canal Zone this insect has been found at- Cristobal, Mount Hope, Gatun, Frijoles, Pedro Miguel, Miraflores, Corozal, Balboa, Palo Seco, and Ancon, and in the Republic of Panama it has been found in Colon, Panama (including the suburbs), the Las Sabanas region north of Panama as far as Rio Bajo, Panama Vieja (Old Panama), and Taboga Island. It was not found at the large citrus plantation at Juan Mina, or at Limon, Summit, Las Cascadas, Empire, Gamboa, Venado, Bracho, Mindi, Toro Point, France Field, or Coco Solo in the Canal Zone, or at Pueblo Neuvo, Matias Hernan- dez, Arraijan, Chorrera, Chepo, Pecora, or Almirante (Bocas del Toro region) in the Republic of Panama, at all of which points in- spections have been made by the authors or by Ignacio Molino. Fig. 4.— Distribution of the blaclc fly in tlie Canal Zone and adjointag parts of the Republic of Panama. The map (fig. 4) shows the distribution of this pest in the Canal Zone and adjoining parts of the Republic of Panama. The hosts on which Aleurocanthus woglumi has been found in various parts of the Old and New Worlds will be found under the heading, ''Food plants," page 14. SPREAD OF THE INSECT IN THE NEW WORLD. Ashby (3) says that Aleurocanthus woglumi "was probably brought here [Jamaica] on mango cuttings from India within the last 20 years." Morrison, after his investigation of conditions in* Jamaica, believes that "the pest has been introduced certainly within 10 years or even as late as 1910." There is no question that the insect was first introduced into Jamaica from India either on mango or some other host, and that from this focus in the New World it has spread to Nassau, Now Providence, Bahama, the Guantanamo and Santiago de Cuba regions of Cuba, and the Canal Zone. The actual time of its introduction into any of these localities has not been determined. THE BLACK FLY OF CITRUS. 7 If it is assumed that it was introduced into Jamaica within the last 10 or 12 years, then its spread from there to the places mentioned before has indeed been remarkably rapid, and it is merely a matter of a short time mitil it will occur throughout Cuba, the Bahamas, and the citrus-growing regions of Central and South America. There is perhaps only one way that an insect like AleurocantTius woglumi can be introduced from an infested region to an uninfested, suitable, far distant one and become established. This is on a weU- infested host plant or parts thereof on which the leaves are allowed to remain, so packed that the various stages of the insect can con- tinue their development throughout the journey, or at least with- stand it. This is because there are several checks to its successful estabHshment: Fu'st, a high mortality of the individuals, especially in the early instars or at molting time, if subjected to adverse con- ditions, such as drying out or heavy rains; secondly, the fact that parthenogenesis occurs in A. woglumi just as in other white flies whose life history has been worked out; and thirdly, the fact that the larvae are not vigorous crawlers and seldom get more than an inch away from the eggs from which they hatch. These factors are con- sidered at length in a discussion of the Hfe history of this insect. When AleurocantTius woglumi is favorably introduced into a new locality, however, there are several ways in which it may spread and become thoroughly established in a region. These methods are the natural and the artificial. The natural method includes: First, the natural flight or migration of the adults from infested to clean plants; secondly, the carriage of the adults and possibly larvae by winds. The artificial method includes: First, the can-ying of infested plants from one place to another either in the form of pot or specimen plants or in numbers as in the case of nursery stock or as cuttings;* secondly, the carrying of adults on vehicles, trains, automobiles, or the clothes of persons passing or working among infested trees. There is no doubt that the black fly has had at least a half dozen or more chances of becoming introduced into the Canal Zone through the shipment of infested food plants from Jamaica. In the Canal Record for July 31, 1912 (6), this statement appears: "Plants and shrubs have been received at Ancon from Director Wilder of the Botanical Gardens at Honolulu, from the director of Hope Gardens, Kingston, Jamaica, and from the Department of Agriculture in Washington, D. C." The Record gives a large list of plants, several of which have been found to be hosts of the insect, but unfortunately no mention is made of the origin of specific plants and the writers have been unable to trace any of the food plants mentioned in the list to Jamaica origin. The important point is that plants have been brought to Ancon from Jamaica as late as 1912 and undoubtedly later, and Ancon, Balboa, and Panama seem to be the most heavily 8 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. and thoroughly infested parts of the Canal Zone and closely adjoining parts of the Republic of Panama. In surveys made by the authors at Ancon, Balboa, and Panama, not a single citrus tree has been found entirely free from tooglumi, and at least 70 per cent of all the mangoes have been found infested, about 10 per cent of these being heavily infested. Practically all the citrus trees in these towns are limes used for ornamental purposes in gardens, in lawns, or along the streets, and fully 50 per cent of such trees are what might be called heavily in- fested. It has also been determined by a survey of the Las Sabanas region, which lies to the north of Panama, that several wealthy Panamanians have themselves introduced plants from Jamaica to that region. There is no regulation of any kind outside of the ordinary customs regulations governing the introduction of plants to either the Canal Zone or the Republic of Panama from any part of the world. Neither are there any. regulations governing the free movement of plants within that area. Hence, it is no uncommon thing to see passengers on vessels from various parts of South America bringing living plants into the Canal Zone or the Republic of Panama, to friends or for their own use. Neither is it an uncommon thing to see passengers from the West Indies or other parts of Central America bringing in plants. These passengers are often residents of the Canal Zone or Repubhc of Panama and bring the plants as a remembrance of their visit or because of the fact that such plants, if they are fruit trees, bear more delicious fruits than do closely related ones in the Canal Zone or Panama. It is common knowledge that the West Indian negroes are great lovers of plants and it has been repeatedly observed by the writers in their visits to Cristobal that even negro workmen carry plants with them when they leave or arrive in the Canal Zone. Prom the foregoing it is apparent that the original infestations of Aleurocanthus woglumi occurring at the terminals of the Panama Canal in the Canal Zone and Republic of Panama may be the result of the introduction of more than one lot of infested host plants from Jamaica, and it is probable that the first introduction took place as early as 1912, Whether the original infestation in the Cristobal-Colon section was a separate introduction or whether it preceded or succeeded the Ancon-Balboa-Panama one will never be known definitely. It must be remembered that the grounds of Ancon Hospital were a sort of botanical garden in the days of the French and remained as such in the early part of the American regime, their place finally being taken by what is now Ancon Nursery. It is quite possible, therefore, that the infested plants might have been sent to Cristobal and Colon from Ancon. THE BLACK FLY OF CITRUS. 9 That this insect has been and is freely spread about on citrus nursery stock is shown by the fact that young lime trees in Ancon Nursery were found to be infested with it. Further numerous cases have been found in Panama where persons have grown young citrus trees of various kinds, mostly oranges and tangerines, in the city from seeds and then taken the plants, when they had grown to be a foot or more high, to their farms in the Las Sabanas region or to Taboga Island, which is 12 miles from the mainland in Panama Bay. Indeed, the heavy infestation on this island can not be accounted for in any other way except that infested food plants of the black fly were brought there from the mainland, and from these as a focus the insect has spread by natural means. Unquestionably the infestations at Mount Hope, Gatun, Pedro Miguel, Corozal, and Palo Seco were started in the same way. At Frijoles and Miraflores the infestations are in all probability "train borne." At the former place lime trees growing by the station and within 25 feet of the railroad track, which were carefully examined in October, 1918, and found to be uninfested, were found slightly infested in April, 1919. In the Canal Zone, the Panama Railroad supplying its commissaries runs its freight cars onto sidings bordered with limes infested with the black fly. This is particu- larly true at Ancon. Hence adults might readily fly onto or into such freight cars and fly off at a place like Frijoles where practically all trains stop. The only trees found infested in the region were five limes by the station ; large oranges and limes around the village, and younger limes at the entrance of the avocado plantation, both of which are at least 150 feet away from the railroad, being free from the pest. The same condition obtained at Miraflores, where two lime trees of a double row of 108 limes leading from the road west of Miraflores Tunnel to the filtration plant were found lightly infested. These trees were the two nearest the raihoad and within 75 feet of it. In traveling on the passenger trains from Panama City to various points along the main line, the writers and Mr. Molino have inspected the windows for the adults of Aleurocanthus woglumi, and although several species of larger insects have been found, at no time has a single adult of woglumi been seen. However, on such trips persons have been seen taking flowers and plants from Colon and Panama to points along the raihoad in the Canal Zone or vice versa. The foregoing method may offer a possibility of the introduction of this insect into Florida from Cuba, as Newell (25) has pointed out. Whether the insects could become successfully established depends on the numbers of females that are introduced and whether or not these females have been fertilized. Its suc- cessful establishment in Florida will depend also on the adapta- 10 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. bility of the insect to climatic conditions that vary considerably! from those in which it is now known to occur. That the spread of this insect in the Ancon-Balboa-Panama, the Cristobal-Colon, the Las Sabanas, and Taboga Island regions is entirely due to the widespread dissemination of infested food plants is not the wi'iters' idea. Doubtless much of the dispersion has been due to this method. But there is sufficient evidence that the infestation does spread to uninfested hosts through the flight or migration of the adults. This has been shown by numerous field- inspection trips. The infestation of large trees such as mango, sapodilla, lime, and orange in these regions, many of them much older than the introduction of woglumi, would show this. At Taboga Island the spread is definitely from the village of Taboga along the shore line to the east and west and to the south following the paths up the steep "backbone" ridge of the island. This spread is most probably a question of the flight of the adults, and the rate at which it is taking place on coffee plants along the creek leading from the village to the south and opposite side of the island substantiates this opinion. That the adults do fly or migrate is shown by the following facts: There is often a decided variation in the number of adults on the young growth in the morning and evening inspections of the same trees, indicating a migration of the adults either at dusk or in the early morning. On the evening of July 29, 1918, Zetek found several adults in his house, indicating that the migration takes place at dusk; adults have been taken in spider webs 75 to 100 feet away from the nearest infested food plant; the migration from an infested food plant to a noninfested one has been demonstrated in Panama; and finally, the observations at Corozal indicate that a flight of 400 feet is possible. The first point mentioned is shown by the seasonal abundance records for a year, observations on three different lots of trees being made three times a day. Adults caught in spider webs were commonly observed in a garden next to the board of health laboratory during the months of May and June, 1919. Like- wise it was found that the adults were flying from heavily infested lime trees to young trees of guayabanon (Annona squamosa) and guava (Psidium guajava) about 60 feet away and set out in nursery rows. Various stages of the pest were later found on the guava, but the annona here remained free from all stages except the adults. In the yard of Ignacio Molino in Panama early in April, owing to a heavy infestation of Aleurocanthus woglumi, two orange trees were "dehorned," every leaf and green shoot being cut off' of them. Early in May, nine new shoots were growing from adventitious buds and on these were found adults of the black fly and egg spirals. A careful survey of aU the plants in the yard was made. At a dis- THE BLACK FLY OF CITKUS. 11 tance of 30 feet from these orange trees a young palm, Eleais melanococca, was found to be lightly infested and adults were seen flying away from it toward the orange trees when distm-bed. Hence, it is evident that this palm was the source of reinfestation of the orange. At Corozal in October, 1918, four lime trees in front of what was once the old post office (house 543, now used as quarters for attend- ants at Corozal Asylum) were found lightly infested with A. woqlumi. Across the road, behind the houses opposite Corozal Asylum, on a long open lawn space, was a row of 29 lime trees, none of them infested, the nearest being at least 300 feeifc from the four infested trees. After the first week of heavy rains following the dry season (Apr. 24, 1919) all trees were reinspected. The infested trees showed a decidedly heavier infestation and of the row of limes uninfested in October, 1918, trees Nos. 9, 10, 20, 25, 27, 28, and 29 were found infested. Tree No. 1 was the farthest away (about 900 feet) from the source of infestation and trees Nos. 20 to 29 the nearest (200 to 300 feet away) to it. Trees Nos. 20 to 25 were almost directly opposite the originally infested trees. That this was a recent infesta- tion was shown by the fact that no individuals were found beyond the pupal stage and no adults had emerged from any of the pupae found. In July this row of trees was again examined and trees Nos. 9, 10, 11, 12, 18, 20, 23, 25, and 29 were found infested and the infestation was increasing. It is evident that the infestation of numbers 11 and 12 came from trees Nos. 9 and 10 and that of 18 came from tree No. 20. The infestation on Nos. 27 and 28 had died out for some reason. Here we have an infestation due directly to the fUght of adults, the longest flight being about 700 feet and the least 200 feet. It has been noted in bringing adults into the laboratory on shoots on which they had gathered in abundance in the field, that when such shoots wilted the adults would leave them and fly out of the screened windows, a case of forced migration guided by positive phototropism. Likewise in working around trees on which adults were clustered in the late afternoon it has been observed that when disturbed many of the adults take to flight and, instead of merely flying off a short distance and returning to the trees as they often do, a large number would disappear permanently. No observations have been made as to whether or not the insect is carried by the wind. Nor has there been seen any such migration as Morrill and Back (21, p. 44-48) have described as taking place in the case of the citrus white fly (Dialeurodes citri Ashmead). This is probably due to the fact that no large and heavily infested citrus 12 BULU5TIN 885, U. S. DEPARTMENT OF AGRICULTURE. groves were available for observation either in the Canal Zone or near-by portions of the Republic of Panama. Infestations started by adults carried on vehicles or on the cloth- ing of men working among infested trees and then going to unin- fested ones have not been seen. However, the first-mentioned wTiter, after making observations on trees on which adults were common, has carried stray individuals on his clothing at least 3,000 feet. The rate of spread or infestation on infested trees has been deter- mined in the following ways: First, taking a shoot that is beginning to grow and watching the number of new spirals laid on the leaves of said shoot; second, taking shoots at the beginning of the rainy season and checking the number of egg spirals laid on the leaves of said shoots. In this latter method shoots of the preceding season were selected and the rate of infestation on this mature growth as well as on new growth that started from the tip of the mature growth was determined by the number of eggs laid on such growth. The first method is illustrated by the record kept on a young shoot of orange from October 23 to November 17, 1918, or from the time that this shoot had grown from 3.5 to 11 inches long. No eggs were laid before November 5 and the leaves are numbered from the bottom of the growth upward. Table I. — Number of egg spirals laid by the black fly on young orange shoot. Leaf No. First spirals laid. Num- ber. New spirals laid on November- - To- tal. 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 11 12 13 Nov. 5 Nov. 6 Nov. 8 ...do.... Nov. 9 Nov. 12 Nov. 13 ...do.... Nov. 14 Nov. 13 2 3 3 2 4 3 2 3 3 0 0 0 0 0 0 0 0 1 3 0 0 0 0 0 0 0 0 1 0 3 2 0 0 0 0 0 0 2 1 10 5 4 0 0 0 0 0 2 1 0 1 0 0 0 0 0 0 2 3 0 2 2 0 0 0 0 0 2 1 0 0 3 3 0 0 0 0 2 2 2 1 1 4 2 1 0 1 0 2 2 1 2 4 1 Nov. 14 Nov. 15 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 No eggs were laid on this shoot between November 15 and Decem- ber 1; observations, therefore, were discontinued. In this case it must be borne in mind that the old growth of this tree was heavily infested with A. woglumi in all stages of development, and that the adults were abundant on the young growth up until November 17, after which a decided dropping off took place, so that during the period from November 23 to December 23 very few adults were present on the entire tree. On April 8, which is just at the end of the dry season, the eggs on a 19-inch long shoot were counted and found THE BLACK FLY OF CITRUS. 13 to be as follows, the count of the leaves being made from the bottom to the top of the shoot. I Table II. — Eggs of black fly on 19-inch orange shoot at end of dry season. LeafNo 1 2 3 4 5 6 7 8 9 10 U 12 13 14 15 16 17 18 19 20 21 22 23 24. Eggspirals 0 0321241015012000000000 0. This shows that the rate of the infestation of young growth on a tree is dependent (1 ) on the degree of infestation of the older leaves on the tree which, in a large measure, determines the number of adults present on the young growth; (2) on the season. This is discussed further undei* the heading, "Seasonal history," page 42. For an idea of this tree at the times these counts were made see the photographs (PL I). That the degree of infestation or rate of spread on a given plant is determined by the species of plant is shown by Tables III and IV for the lime trees at Ancon. Table III. — New egg spirals of the black fly on old and new lime leaves, experiment No. 8, 1919. Leaf No. May 17. May 24. May 31. June 7. June 14. June 21. June 28. July 5. 1 0 1 0 0 0 0 1 2 0 0 0 0 0 0 1 0 0 0 1 2 1 0 0 0 0 0 1 2 2 0 3 0 0 3 0 1 1 1 0 3 1 0 1 0 1 0 2 1 0 0 1 1 1 1 0 1 3 0 1 2 0 0 1 0 0 0 0 1 0 (•03 2 1 0 2 0 0 0 0 3 0 0 0 4 0 0 0 0 0 0 0 0 2 0 0 0 0 0 2 3 4 5 6 7 8 Ai 2 0 B C D 0 E Total 2 5 14 10 11 7 7 7 » The leaves on the new growth were lettered A, B, etc., upward from the junction of the new growth with the old and the leaves of the old growth were numbered I, 2, etc., upward, starting with the junction of a much older shoot or branch. The observations on this shoot were begun on April 10, 1919, at the end of a dry season, the ramy season beginning on April 13. Table IV. — New egg spirals of the black fly on old and new linne leaves, experiment 9, 1919. Leaf No.i Apr. 17. Apr. 26. Mays. May 10. May 17. May 24. May 31. June 7. June 14. June 21. June27.» 1 1 2 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 1 0 1 2 2 1 0 1 1 0 3 0 0 0 0 0 0 0 1 2 0 0 0 1 0 0 0 0 0 0 5 5 2 0 7 1 1 1 1 2 1 0 0 0 0 1 3 A» 0 4 0 0 0 0 0 0 2 1 1 0 ; 1 0 0 0 0 1 B C D E 0 2 F G 1 H 1 I I J.... 0 1 In this experiment the lettering of the young growth and the numbering of the old are the same as in thepreceding. ' This experiment was accidentally destroyed by the pruning of the trees on June 30. • This shoot started to grow on May 3, but no eggs were observed on it until May 24. 14 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. These 2 tables are characteristic of 30 that were kept on lime trees. The field observations at Ancon, Balboa, and Cristobal indi- cate that an infestation of Aleurocanthus woglumi on this host, while persistent and oftentimes heavy, does not progress with the rapidity that it does on such hosts as grapefruit, orange, and tangerine. Further discussion of this point is given imder ''Injury," page 18. In this connection see Plates I, II, and III. Many of the trees along the roads and on the lawns in Ancon came to the Canal Zone in 1916 as nursery stock from Aguadulce, Republic of Panama, and all those trees are more or less uniformly infested, indicating either that they were all infested at the time that they were set out, or that they became so shortly thereafter through eggs laid on them by migrating adults. It has recently (Sept. 15, 1919) been ascertained, however, through a letter accompanied by specimens, that Aleurocanthus woglumi is firmly established at Agua- dulce. This would indicate that the trees were infested at the time they were set out, and if the rate at which the infestation on the trees is progressing is any criterion by which to judge the time that woglumi has been present in the Canal Zone, it is safe to say that it was fu-st introduced there not later than 1912, for these lime trees must yet, after three years, be regarded as only moderately infested. On this basis and assuming that it came to the Canal Zone dinging or before 1912, then it has been in Jamaica 15 or more years or at least since before 1910. In Jamaica and Cuba the insect's spread has paralleled that in the Canal Zone. In the former place after being introduced into the vicinity of Kingston it spread around this locality by both natiu*al and artificial means, but its dispersion to the more distant parts of the island was unquestionably by means of infested food plants. Once it was introduced into a new locality its spread there was by both natural and artificial means. In Cuba it was first introduced into the Guantanamo region on infested plants from Jamaica and its spread took place by both natural and artificial means. From Guan- tanamo it was carried to Havana on infested plants, where its spread was again by natural and artificial means. Whether or not the in- festation in the region of Santiago de la Cuba is a separate introduc- tion of infested plants from Jamaica or whether it is traceable to infested plants from the Guantanamo region is not known. FOOD PLANTS. The food plants of the black fly fall under the three following heads: (1) The favorite or preferred food plants, i. e., those that be- come heavily infested with the insect and on which complete develop- ment from egg to adult takes place; (2) the occasional food plants, i. e., those on which complete development of the insect can and does THE BLACK FLY OF CITRUS. 15 take place, but which apparently do not become heavily infested with it; (3) supplemental food plants, or those which the adults may visit and from which they may obtain food, but on which either they do not lay eggs or, if they do, complete development does not take place. It must be borne in mind that these three classes are not clear-cut ones but merge into one another. They are based on the winters' observations in the Canal Zone. In this region the favorite food plants in the order named appear to be Ardisia revoluta, the oranges (both sweet and sour) , grapefruit, lemon, lime, and mango. No large areas of coffee in infested regions have been foimd in our surveys and little can be said about this host from oiu" experience, though the writers in Cuba and Jamaica are agreed that it comes before mango in the list. Among the occasional hosts may be included such plants as sapo- dilla (Achras sapota), oil nut palm (Eleais melanococca) , Cashew apple (Anacardium occidentale), sugar apple (Annona squamosa), Eugenia malaccensis , mamon (Melicocca hijuga), guava (Psidium guajava), and mamei (Lucuma mammosa) . With the exception of the favorite or preferred food plants these constitute an important means by which the insect may be spread on niu^ery stock or on individual plants. These occasional food plants may and probably will con- stitute an important source of reinfestation where control of the black fly, especially on citrus, is undertaken. This has been shown in the case of the oil nut palm (Eleais melanococca) under the headmg "Spread of the insect," on pages 10-11. Among the supplemental food plants are the orange jessamine {Chalcas exotica), Barbados cherry {MalpigMa glabra), and crape myrtle (Lagerstroemia indica). These may serve merely as occa- sional food plants for the adults or as congregating places, but why the adults should congregate on them in large numbers, even in the presence of their favorite food plants, and not lay eggs, as has been observed in the case of the last two mentioned plants, we will not attempt to explain. That it does not always pay to ''jump at con- clusions" or to make broad generalizations regarding the food plants of a polyphagous insect like the black fly is shown m the case of the orange jessamine. In the first place, it is a rutaceous plant, a citrus relative, and one would naturally expect it to fall into fii*st or second class of food plants. That it is merely a supplemental host has been proved, not only by the repeated examination of a large number of plants, but also by attempting to rear the insect on it. Never has a single specimen of Aleurocanthus woglumi beyond the fu'st instar been found on this plant, though egg spirals are laid in abundance on it and adults often collect in numbers on its younger growth. Although the larvae are able to attach themselves to the leaves, mvariably they die and fall off before reaching the first molt. Hence, 16 BULLETIN 885, U. S. DEPAETMENT OF AGRICULTURE. simply because the adults of woglumi gather on a plant or because eggs and the early stage larvsB have been found on it, the conclusion that such a plant is an important host of the insect does not neces- sarily follow. The only criterion as to whether a certam species is in reality a favorite or an occasional food plant is the finding of insects in all stages of development on it. Aleurocanihus woglumi Ashby has been recorded from 75 different food plants or hosts. These are distributed among 30 plant families. The following table shows the plants on which it has been found, the family to which the plants belong, and the place and person who collected it. The plants on which it has been found in the Canal Zone and the adjoining parts of the Republic of Panama are marked with an asterisk and a number, the number indicating the class of the food plant as previously discussed. The writers do not presume to say to which class the food plants recorded by other authors belong, but such plants as avocado, hibiscus, various begonias, papaya, croton, plantain, pomegranate, star apple, and coral vine have been repeatedly examined m the Canal Zone and Republic of Panama, and in no case have any stages of the black fly been found on any of them so far. The banana is an example of a plant that is on the border line between the second and thu'd classes. Out of at least a hundred plants that have been carefully examined only three colonies of the insect have ever been found on this host, and those were on a plant growing within 6 feet of a mango heavily infested with the insect. Table V. — Host plants of the black fly {Aleurocanihus woglumi). Scientific botanical name.i Acalupha linncolataf {lan- ccolata Willd). Achras sapofa Linn. (x-2).. . Anacardium occidentale Linn. (x-2). AnnoTM cherimola Mill Annona muricaia Linn Annona squamosa Linn. (x-2). Annona sp. (x-2) Antigonon leptopus Hook . . . Ardisia re valuta H. B. & K. (x-1). Bassia latifolia Roxb Begonia sp Capparis pedunculosa Wall. Capparis roxburgM DC C'arica papaya Linn Family. Euphorbiaceae. Sapotaceae Common name (Spanish and English). acalypha {nispero, sapote . sapodilla f maranon >Anacardiaceae... cashew apple Annonaceae !|chirimoya '\Jamaica apple. do i/anon manteca soursop » lanon |....do do {rosa de montana, coral- ito. coral viae, love's chain. Myrsinaceae I f ruta de pava ...^ mahwa. e I begonia. Capparidaceae... do caper-bush do ffruta bomba, papaya ipapaya Locality and by whom reported.' Guantanamo, Cuba. — C. J. & H. Canal Zone, Panama, R. P. — D. &Z. Cuba (Habana and Guanta- namo).— C. J. & H. Guaatanamo, Cuba. — C. J. & H. Las Sabanas, R. P.— M. & Z. \Guantanamo, Cuba. — C. J. & I H. } Do. Guantanamo, Cuba.— C. J. & H. Las Sabanas, R. P.— M. & Z. Canal Zone.— D. & M. [Guantanamo, Cuba. — C. J. & 1 ^- Las Sabanas, R. P.— M. & Z. Jamaica. — R. Guantanamo, Cuba.— C. J. & H. Royal Botanical Garden. Cey- lon.—W. Q. & B. Do. \ Guantanamo, Cuba. — C. J. & / H. For footno tes 1 and 2 see page IS. THE BLACK FLY OF CITRUS. 17 Table V. — Host plants of the black fly (Aleurocanthus woglumi) — Continued. Scientific botanical name. Family. Common name (Spanish and English). LocaUty and by whom reported.' Ccstrum diurnum Linn Ccstrum nocturnum Linn . . . Chalcas exotica Millsp. (x-3) . Chrysophyllum cainito Linn . Citrus auranlifolia Swingle {limelta anct.) (x-1). Citrus aurantium Linn. (x-1) Citrus grandis Osbeck {dc- cumana Linn.) (x-1). Citrtis limonia Osbeck (x-1) , Solanaceae . ....do.-. Rutaceae . Sapotaceae. Rutaceae. . .do. ...do. .do. Citrus medica Linn Citrus Ttobilis dcliciosa Swin- gle (x-1). Citrus sinensis Osbeck (x-1) . Citrus spp ..do. ..do- ..do. ..do. Citrus sp. not specified. .do. Citrus sp. not specified Clausena lansium Skeels ( Clausena wampi 0!iv.). Coffea arabica Linn, (x-1) . . Cordia alba Roem. and Schult Cordia sp Crescentia cujete Linn Croton sp., probably Co- diaeum referred to. Cupania cubensis {BligMa sapida Kon.) Eleais melanococca Gaertn. (x-2). Eugenia jambos Linn. (x-2). Eugenia malaccensis Linn. (x-1 or 2). GtKiicum officinale Linn Cfuazuma tomentosa Knuth . Hibiscus rosa-cMnensis Linn. Hibiscus schizopetalus Hook . Iiora thwaitesii Hook (x-3) , Kurrimia ceylanica Am .do. .do. Rubiaceae . Myrtaceae. k...do fgaian de dia Viay-blooming jessamine. [galan de noche -,T r - ^ $ •~ y # ... ^ ^ • I' 0 *-* 9 b' ■** Stages in the Life History of the Black Fly. A, Eggs and first and second instar larvee; B, eggs and third instar larvae; C, male pupae; D, female pupae. All enlarged to the same scale (taken with 48 mm. lens). THE BLACK FLY OF CITHUS. 21 but the injury seemed to be due to neglect and to the fungus (Gloeo- sporium gloeosporoides) , which had gained entrance to the trees through pruning wounds. At Corozal the injury seemed to be due to heavy infestations of the purple scale (LepidosapJies heckii Newm.), and of the West Indian red scale [Pseudaonidia (Selenaspi- dus) articulatus Morg.]. These dying lime trees were not heavily infested with A. woglumi. In the authors' work at Ancon, Balboa, and Cristobal, it was foimd that when citrus trees Gime, orange, and grapefruit) were heavily in- fested with the two scales mentioned above, they were almost invaria- bly only lightly infested with A . woglumi even though they were in close proximity to trees heavily infested with this insect. From the authors' experience then it would seem that the infestations of scale insects and the black fly go on independently of each other and it does not neces- sarily follow " that an infestation of the black fly is invariably accom- panied by rapidly increasing scale infestations so that the life of an infested tree is necessarily short" (25). That this may take place is not denied, but that it has not taken place in the Canal Zone to date is the result of a year's observation in that region. It has also been observed that heavy infestations, especially of the West Indian red scale, serve as a decided inhibition to the development of the black fly, as will be pointed out under the life history of the insect. The writers fiud that the two scales mentioned do far more actual and noticeable damage than the black fly in that they actually kill infested areas on the leaves and cause infested twigs to die. In many instances the cultural conditions and neglect to which trees have been subjected are in themselves sufficient to injure them seriously even if no insect or fungous pests were involved and such factors coupled with a heavy infestation of the black fly doubtless will, in time, prove a handicap that these trees will be miable to over- come in spite of the fact that the chmatological conditions of this region are extremely favorable to plant development. Since in all the survey work conducted by the authors a tree was never found killed by the black fly, it is safe to conclude that much of its injury is "ornamental" and that the fact that it makes heavily infested trees decidedly misightly has been largely responsible for the exaggerated statements regarding its destructiveness. In conclusion the authors' work bears out the following statements of Morrison, who has seen the insect and its injury in the Canal Zone, Jamaica, and Cuba: "It may be said that the presence of this pest in numbers affects adversely the infested tree, and may in cases where other factors are also unfavorable be the final* handicap which pre- vents the production of a crop of fruit or reduces it in size imtil it is unprofitable commercially to grow such a crop, or prevents the trees from making the amount of new growth each year which is necessary 22 BULLJiJTIN 885, U. S. DEPARTMENT OF AGRICULTURE. to their proper development. There is nothing to indicate that, if it succeeded in entering the United States, it would prove to be an almost hopeless case * * * though it would without question prove to be a heavy additional load for citrus growers already bur- dened with numerous other injurious insects and diseases." LIFE HISTORY AND HABITS. In this work certain difficulties due to the habits of the adults of A. woglumi had to be overcome. These habits made it impossible, in many instances, to secure much desired data by the direct method. The data desired were the number of eggs laid by a single female; whether unfertilized females lay only eggs that give rise to males; how soon after emergence copulation takes place; the length of life of the adults; and how soon after emergence egg-laying begins. The difficulty of finding out these facts by the direct method, namely, by taking individuals of known history and confining them with leaves in cages (PL VII) or in petri dishes, was due to the nervous- ness of the adults when they were disturbed or handled. Whether adults were confined singly or in numbers seemed to make little difference. Once they were disturbed and confined, they would worry themselves to death in their attempts to escape and instead of resting on the leaves of the plant to which they were confined the)^ would wander up and down the sides of the cage. Often pupae from which adults were ready to emerge were placed in petri dishes with fresh leaves, but almost invariably on emergence the adults, after becoming thoroughly colored, would begin wandering around the cage in their attempts to get away. In no case did adults obtained in this way live over four days. Therefore, another and less accurate method of observing the habits of the adults and obtaimng spirals for life-history work had to be adopted. Large numbers of males and females were brought into the laboratory on the young shoots on wliich they were found resting in the field. Copulation was usually in progress and this insured obtaining normal fertilized eggs. These shoots were placed in water and as they wilted the males and females would leave them. Therefore, such shoots were set among young trees so that the adults could congregate on these. For a time this method was successful, but it often failed, apparently due to the fact that the females preferred to lay eggs on the wilted leaves rather than on those of the trees that were provided. This may have been due to the age of the leaves on the trees, but no such pronounced selection on the part of the females in the field has been observed. In the field, spirals which from their color and from actual observation were Imown to have just been laid were used. Those were chosen that occurred alone on the leaves, or if any other spirals happened to be laid on the same leaf, which was rarely the case, THE BLACK FLY OF CITRUS. 23 these were removed. The chosen spirals were observed daily, or three times daily in many instances, imtil the life cycle was com- plete. Hence by close observations in the field and by a careful study of the habits of the insects in all stages in the laboratory, answers to most of the questions mentioned before, and which could not be determined by the direct method, have been obtained. Two methods were used in obtaining the life history of this insect, namely, the group method and the individual method. Both had certain advantages and disadvantages. The gi'oup method con- sisted in taking an egg spiral laid on a known date, watchuig for the eggg'to hatch, and then making observations on the individuals from one to three times daily until they had completed then* life cycle. In this case, however, only the number of individuals that passed from one instar to another on a given date was recorded day by day. The individual method consisted in accm-ately plotting the egg spiral and the position of the larvae after they had settled and giving each one a number to be retained by it throughout its development. The time when each molted was then recorded, thus giving a very accurate record of the development of each individual. The advantage of the group method is that it is a rapid one and shows readily by a glance at the chart how many individuals die in a given instar, the duration of a given molt, and the time of the maximum molting for a given molt. (See fig. 5.) From such a chart the life-history curve of the individuals of one lot of eggs can be plotted directly. The gi'eat objection to this method is that it does not give a clear-cut individual record, which is also desirable. The advantages and disadvantages of the individual records are just the inverse of the group record, but the chief objection in the work here was that this method was too slow and there was in many cases a waste of time due to the great mortality of A. woglumi, especially in the first instar, as v»^ill be shown later. The two kinds of records taken together make an ideal com- bination, as one can be used to interpret the other. The following tables are representative of the two methods of recording the life history and show the range and variation under favorable (field) conditions and more or less unfavorable (laboratory) conditions. In the laboratory the small trees used were watered daily, but the hxunidity was often as much as 20 per cent lower in- doors than out-of-doors, and there was generally a considerable breeze blowing through the room. Likewise the light in the room was dull and it seems that all the factors taken together retarded the indi- viduals, especially in the pupal stage. But laboratory data give a clue to the ability of the species to adapt itself to its environment and likewise indicate what takes place in the dry season, especially in a region like Las Sabanas. 24 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE, It is seen, therefore, that Ashby (3) is right fundamentally re- garding the life history of the black fly. But he did not go far enough in his studies. The life history is not clear-cut, as one is led to believe from his brief work, and is greatly influenced by con- ditions. The total length of time from the egg to the adult as given by him and Cardin (11) may be much longer than they indicate. Some of the individuals do go through their development in as few as 55 days, but there is a great variation and the majority take a Fig. 5.— Diagrammatic chart showing method used in keeping record of individuals of Aleurocanthus woglumi in life-history work. i . longer time, especially under unfavorable conditions. There is con- siderable variation in the length of time that individuals remain in each instar, the greatest variation being in the time that they remain in the pupal stage. This is true in a colony all of whose individuals j j hatched from the same egg spiral. There seems to be a constant . dropping back of some individuals so that the group life history is drawn out. The time for the complete development from egg to adult, as is shown by the tables, ranges from 55 (45 the minimum) to 113 (the maximum) days. THE BLACK FLY OF CITRUS. Table VI. — Life history of Aleurocanthus woglumi. GROUP RECORD 2 (1918), INDOOR EXPERIMENT. HOST: LIME. 25 Eggs laid. No. Hatched. No. First molt. No. Second molt. No. Third molt. No. Adults. No. Sex. I M.i F. Aug. 6« 55 55 Aug. 17 Aug. 18 Aug. 19 30 20 5 55 Aug. 25 Aug. 263 30 17 Sept. 1 Sept. 2 Sept. 3 Sept. V 6 9 17. 2 34 Sept. 10 Sept. 11 Sept. 12 Sept. 13 » Six second larval instar indi\iduals died in this stage. M Eggs all hatched; two first larval instar indi\T duals fell off before settling down. »* Two first larvalinstar individuals fell off in molting. IS All eggs hatched; 4 first larval instar individuals fell off before settling down. w See 1', 4 second larval instar individuals washed off. 1' On June 5 and 6, 31 first larvalinstar individuals and 4 second larval instar individuals were washed off by heavy rains at molting time. 26 BULLETIN 885, U. S. DEPARTMENT OF AGRICULTURE. Table VI. — Life history of Aleurocanthus woglumi — Continued. GROUP RECORD 42 (1919L OUT-OF-DOORS EXPERIMENT. HOST: ORANGE. Eggs laid. No. Hatched. No. First molt. No. Second molt. No. Third molt. No. Adults. No. Sex. M, F. May 15 32 32 May 27 May 28 May 29 May 30 May 31 June 1 w 6 0 3 0 10 12 31 June 3 Jime 4 Jvme 5 June 6 June 7 June 8 19 5 5 6 2 6 5 29 Jime 10 June 11 June 12 June 13 June 14 June 15 June 16 June 17- June 18 June 19 20 1 1 2 8 2 3 6 0 1 24 Jime 19 Jime 20 June 21 June 22 Jime 23 June 24- June 29 June 30 July 1 July 2-4 July 521 1 0 2 4 2 5 2 2 0 1 19 July 12 July 14 July 15 July 16 July 17 July 18 July 21 July 22 July 23 July 2522 3 14 3 1 1 .... 6 Total "i 3 1 1 "i 1 8 GROUP RECORD 47 (1919), INDOOR EXPERIMENT. HOST: LIME. May 10 81 May 21 May 22 4 17 May 30 May 31 5 6 June 5 June 6 3 5 June 14 June 15 1 5 July 11 July 14 1 1 May 23 28 June 1 5 Jime 7 4 June 16 3 July 17 3 1 May 24 5 June 223 tt June 8 5 June 17 2 July 18 1 1 May 2521 4 June 323 12 June 9 1 June 18 2 July 20 1 1 June 4 6 June 10 5 Jime 19 4 July 25 1 1 June 525 5 June 11 1 June 20 0 July 28 1 1 June 12 6 June 21 1 July 29 1 1 June 1326 7 June 22 June 23 June 24 June 25 June 262' 7 2 1 2 3 July 31 Aug. 1 Aug. 2 Aug. 4 Aug. 10 Aug. 15 Aug. 18 Aug. 20 Aug. 22 Sept. 3 Sept. 19 Sept. 25 Sept. 28 Sept. 3028 1 2 2 2 1 1 1 2 1 1 1 2 1 2 1 2 1 2 "i 1 '"2 1 2 Total 81 58 45 37 33 30 11 19 18 All eggs hatched; 1 first larval instar individual lost in settling down. 19 Two first larval instar individuals died in that stage. 20 Five second larval instar individuals died in that stage. 21 Five third larval instar indi\'iduals died in that stage. 22 Tree sprayed with nicotine oleate on July 25 and 5 pupse, 3 females, and 2 males were killed. 23 Three second larval instar indi\'lduals removed for study. These had just molted. 2< Fifteen eggs failed to hatch. These were not parasitized. Eight first larval instar individuals were lost before settling down. 25 Three first larval instar individuals (still alive) removed for study. Ten first larval instar individuals died in this stage. 26 Two second larval instar individuals died in this stage. 27 Four third larval instar individuals died iu this stage. >8 Three pupae still unreported on. THE BLACK FLY OF CITRUS. 27 Q s 2 S* o o o ■,< « I Q ^ 5 m ■< H« • a S^- N • e ^ <^ : iNlHi a i 3 . . . . . ft . • ■ ft • • ^ • ft • ■ ■ ~ ; !o ' «r; ; ;oo ; :(N I ; bi I I bi . . 3 . . a 3 • £ ; :^ : :^ -^ : t 1.^^ • '^^ V— / ' w s . 1 .O • 'OS 05 '• a e e C) "o ; ; :a i :a a : a : : ; 03 : ; 03 . 03 I 3 i Mt i it Id 1 .;;!-,. .hr. ! I . r* . .05 •t^ • a 5^ a H O '10-* •■<)< ; I"" •* : 1 o a >> ' ' ! I 1 I 1^ -i 3 dodddoo a o S^.'Tft'^ : ft'^cs ; a 2 'a-d'O'O'O'O'O'o::;:; ►^^'tJ t>>-j>^ ; | f i, :;:::;;( Z^fJnfJHH^pM !l-sf!^l-3 ; 6 . HP) • ■ HM HN Hffl • : 00 t^ t^ t^ o> t^ t^ • • t^ 00 t^ t^ t^ OO 00 ■ a e tH Ci M ; 1 : I ; ; 1 9 :::::;; 1 1 : '; 1 '; j ; jcD 3 aa::aaa::aasaaaa| a fl'^oo"!- '• loTt-Tt-r ; ;[-"oot-t-t-oooo53- (2 pi. -M c^co-*moi^oooso^j^M^2S^2 , 00 3 ® 1^ a t3 iS w B 1-5 , 1 OJ i^ '6 a M ao W a i-j CO • -co .rt Oe^,-, . ,,-1 : >;.'^ >.>> :'§>;. la '■'B'B '.-^3 •l-S •Hsl-5 .Hl-5 ■ r>- 1^ OS t^ • .-H i>- H : : &5 ;ga ; iaaad i^a ; c3 ft : • fto3 ftft HH5(-,0 ■h-5l->l->>^ 1-1 1~ -Hi 1-9 N I:- Oi O Ci i-H i-l O 05 ;"^2 tUD aa isa^aa i^ _ fto5;ftft.o3o3;". 03 g ^ ~03 .^ ; 03 _-. ^ t^Oi • 00 O ^O 00 . „ _ D 3 3 : 3 3.3.3 ^^33 3rN'3 -■ ^a 3 tM n CO ■* lO ?o t~ OC 05 O ^ -i i? H &i CQ o O ^ M CO s en ^ « S n u O p ft S '^ ft •S tf r«; o '^ P4 s tf 1 hJ t-J ft P4 > n ft •259: is a a C3 P. aa "39 a«5aa Sa^Sg^ 5ft>|oOh?5ft*l^^|.l5o p.S'O cj 03 aa ■«1ai LO'TfTo" 05 10 05 00'H aaaa ^ ^ C3 ^ o'io'tiTic' bib &£ bib be S 3 3 3 aaaa bcbf> 3 3 3 3 a>?aaaa^ . 3 .... 3 P,Hj p.p,D.o3i-5 t^laf^ >. >. >;, >^;:3 " >> 3,^3333,=" 13 >-3 ft l-S 1-5 t-5 H, fi* .1-5 0500O030500O000000 aaa^a aa a : a^ '-' o £?aa 3 • • Hr> PiP< p■.-: >^ >> >. t>. 3 >>>> mi-Sl-5 -^INeO'!}iiOf I Bui. 886, U. S. Dept. of Agriculture. Plate Ml. OVI POSITION SCARS AND LARV/E OF SPOTTED ApPLE-TrEE BORER. A , Long chain of punctures in apple bark accompanying oviposition scar. Eggs were deposited at broad place in chain slightly above the center. B, Short chain of punctures accompanying oviposition scar. C, Other forms of oviposition scars. D, E, Pairs of young larvEe found in oviposition scars. Bui. 886, U. S. Dept. of Agriculture Plate IV. jj. iisu -tm ^ t 1! Work of Spotted Apple-Tree Borer. A Nearly mature larva working in heart of apple wood; natural size. B, Pupa occupying natural position at upper end of gallery in heart of apple wood; slightly enlarged. C, Section of trunk of young apple tree having a swollen and roughened appearance as a result of mjury by borers. Bui. 886, U. S. Dept. of Agriculture. Plate V. Work of Spotted Apple-Tree Borer. .4, Wound of borer and exit hole of adult; natural size. B, Wound made by woodpeckers in removing borer from apple wood. C, Exit holes of adults; natural size. D, Apple twigs showing Dark gnawed off bv beetles. SPOTTED APPLE-TREE BORER. 5 in width at the widest point. Eggs in normal position are sur- rounded by a waxy semitransparent substance, which exchides air and probably shuts out natural eneniies. The eggs are placed between the bark and wood at the side of vertical slits in the bark which the f etoale beetle makes with her j aws. (PI. Ill, A, B, C.) These slits, or oviposition scars, may be along the central or upper portions of the trunk or in the branches. They are often accompanied by a chain of shallow punctures extending above and below the oviposition scar, the line of marks resembling the im- pression that would be made by pressing the cutting edge of a fine- toothed saw against the bark. (PI. Ill, A.) In some cases only one egg is placed in a slit and in other cases a pair are deposited, one on each side of the opening. The eggs occupy an oblique position to the line of punctures and are thrust back under the bark from 2 to 3 mm. from the slit at the nearest point. (PI. II, E.) Four eggs laid on June 29 were kept under observation until July 20 and were then lost by accident,, apparently as they were just ready to hatch. This indicates a period of incubation of something over three weeks. THE LARVA. The larva (PI. I, B; PL III, D, E; PL IV, A) is a whitish, deeply segmented, footless, sparsely bristled grub, having a brown head with black jaws and a broad, convex, papillose, shield-like plate on the dorsum of the prothorax. Full-grown larvae are from 25 to 30 mjn. in length and about 5 mm. in width at tlie prothorax. The prothoracic segment is widest, the segments tapering to the fourth, and from there being uniform in width to the twelfth. The thirteenth segment narrows to a blunt point. To the casual observer the larva of this species is indistinguishable from that of Saperda Candida. In the latitude of West Virginia the larvae hatch and begin feeding at any time from the latter part of July until some time in September. During the first season the larval excavations are in the form of more or less circular or roughly elongate burrows on one or both sides (according to whether there are one or two larvae present) of the oviposition scar. (PL III, D, E.) At first fine, sawdust-like castings issuing from the oviposition seal's mark the presence of the feeding borers. The second season the borer enters the wood, forming a ragged, winding burrow that leads ultimately to the heart. Here, in the center of the wood, it mines, usually going upward 5 or 6 inches, where it finally pupates at the upper end of the gallery, after having terminated its burrow with an abrupt curve outward to the inner bark. (PL I, E, F; PL IV, A, B.) As the borer feeds it ejects part of its castings through the bark and packs the rest in unoccupied corners of its burrow. The burrow below the pupal chamber is 6 BULLETIlSr 886, U. S. DEPARTMENT OF AGRICULTURE. packed for an inch or two with excelsior-like strings of wood (PL IV, B), evidently as a safeguard against the accumulation of too much moisture while the insect is undergohig its transformation to the adult stage. Probably the behavior and the metamorphic development of this species are affected to a considerable extent by differences in latitude. No doubt the borer feeds more extensively and grows more rapidly the first season in the South than in the North. In West Virginia, of 16 individuals in which the duration of the larval period was ob- served, 1 lived in the tree 2 years, 14 lived 3 years, and 1 lived 4 years. THE PUPA. The pupa (PI. I, C, E, F; PI. IV, B) is characteristic of the family, being creamy white in color with the rudimentary legs, wing pads, and antennfe folded close to the body. In form it is slender, averag- ing 1 7 min. in length by 5 m.m. in width. The dionensions of the cell which it occupies (PI. I, E, F; PI. IV, B) are only slightly greater than those of the pupa itself. The average size of this cell is about 20 by 8 mm. The restricted quarters of the pupa limit its move- ments much more than is the case with the pupa of the common borer, Saperda Candida, whose pupal cell may be 2 inches or more in length. In West Virginia pupation takes place in April, one speci- men being found in the pupa stage on April 3. The pupal period lasts from four to six weeks. This stage, also, of the insect's develop- ment is probably affected in time liinits by differences in climatic conditions due to latitude. THE ADULT. The adult of this borer (PI. II, A-D) is a handsome beetle, the females of which are from 15 to 20 mm. and the males from 12 to 16 mm. in length, the females being much more robust than the males. The upper parts are cinnamon brown with a broad white stripe on each side of the thorax, a large, oblong, white spot, notched at the ends, on the middle of each elytron, and a smaller, comma-shaped, white spot midway between each of these and the apex. There are usually a minute white spot on the thorax just in front of the scutel- lum and one each on the humeral angles. The sides are white and the underparts of the abdomen and thorax, and the head, antennae, and legs are brown. After transforming, the beetles remain in their pupal quarters for several days to harden and then escape by gnawmg a circular hole through the bark barrier that has shut them in. (PL V, A, C.) Th§ exit hole in the bark is from 5 to 6 mm. in diameter, that of the male being smaller than that of the female. On escaping, the beetles seek seclusion among the branches where they rest in comparative inactivity during the brighter part of the SPOTTED APPLE-TREE BOPvER. 7 day. Near evening they become more active, flying, feeding, copu- lation, and oviposition being engaged in most freely about sundown, although all these activities extend into the night. The beetles feed rather freely on the tender bark, a favorite place for feeding being the thickened bark around the base of small twigs. TIME OF EMERGENCE OF BEETLES. According to Osborne/ the beetles appear in Iowa about the middle of June. Wolcott states, in correspondence with the writer, that the beetles have been taken in Illinois from June 12 to July 20. In 1916 the writer found that apparently the last of the beetles had emerged at East Lansing, Mich., on June 26. On June 15, 1917, however, emergence had not yet begun in the same locality. At French Creek, W. Va., beetles issued in 1919 from May 14 to 31, as indicated in Table I. Table I. ■Time of emergence of beetles of the spotted apple-tree borer at French Creel:, W. Va., in 1919. Date. Number of beetles. Males. Females. Total. May 14 1 1 2 1 1 1 1 2 1 0 0 0 0 0 0 0 0 0 0 1 ^15 1 19 20 1 22 ] 24 1 25 1 26 q 27 r 28 1 29 ] 31 ] Total 11 5 1£ It appears from the foregoing data that emergence may take place from the first of May to the last of June, according to locality and climatic conditions. It is very probable that the beetles appear before the first of May in the more southerly localities of the species' range. OVIPOSITION. The interesting process of oviposition begins a week or two after the eggs are fertilized. As has been stated, the female usually selects the upper part of the trunk of young trees or branches an inch or two in diameter in which to place her eggs. Occasionally the ovi- position scar is in the form of a small, somewhat circular opening in the bark, but more frequently it is a deep, narrow gash in the bark, often accompanied with a more or less prolonged line of pricks ex- tending vertically from one or both ends of the gash. (PI. Ill, A-C.) 1 Osborne, Henry, op. cit. 8 BULLETIN 886, U. S. DEPARTMENT OF AGRICULTURE. Ovipositioii was observed repeatedly and the following may be given as a typical example of one of the more elaborate operations attending the act: About an hour before sundown a beetle was ob- served to take a horizontal position across the trunk of a young apple tree, about 30 inches above the ground. She began at once to force her jaws their full length into the tender bark. (PI. II, C.) After the jaws were inserted and withdrawn, the beetle moved sidewise up the trmik for a short distance and again inserted them and thus continued the operation until she had a straight line of pricks about 2 mm. apart. When the line was about 2 inches long she joined several of the pricks together with a deeper furrow and then turned around, as though on a pivot, and inserted the tip of her abdomen into the opening. (PI. II, D.) She then spent nine minutes in forcing her ovipositor forward between the bark and wood, after which she withdrew her ovipositor, reversed ends, -moved across to the other side of the opening and then consumed an equal i3eriod of time m again inserting her ovipositor. The two eggs having been placed on opposite sides of the opening, the beetle then took her original position and continued making the chain of pricks until the line was 5 mches in length. She then crawled back among the branches of the tree. When the oviposition scar was examined closely the two openings where the ovipositor had been thrust in were visible, both being filled with a brownish, semitransparent substance resembling wax. Occasionally a series of punctures is found in which there occur two of the oviposition furrows, each having one or two eggs. LENGTH OF LIFE OF BEETLES. Most of the beetles kept for observation were confined in roomy wire-screen cages placed over young apple trees in the field. In this position their length of life was probably something near normal. About one-half of the beetles died within a month after emergence, but a few lived for a considerably longer period. One male attained an adult age of 52 days, while a female lived 93 days. The latter individual died on August 30 and continued to oviposit up to within a few days of its death. NATURAL ENEMIES. By far the most effective natural check to the increase of this borer seems to be the woodpeckers. The borers feed in positions easily accessible to these birds and empty buiTows are to be found on almost every infested tree, with the marks of the birds around the wounds giving unmistakable evidence of the cause of the borer's disappearance. (PI. V, B.) During the present studies every attempt to rear larvae in unprotected trees met with a loss of all the individuals as a result of woodpecker attack. The species of bird SPOTTED APPLE-TREE BOllEK. 9 responsible for the loss of the borers was not determined definitely, but all the evidence pointed to the downy woodpecker, Dryohates imhescens medianus. It seems probable that the spotted apple-tree borer would be a much more widely known and destructive pest were it not for the constant depletion of their numbers by wood- peckers. On one occasion a newly emerged female beetle was found in one of the rearing cages hanging suspended in the jaws of a spider. (PL I, D.) The spider was captured and was determined later by Mr. C. R. Shoemaker as Xysticus ferox (Hentz). In several cases in Michigan larvae of the clearwing moth Aegeria 'pyri Harris were found as the sole occupants of burrows made re- cently by spotted apple-tree borers. There was good reason to believe that the larvae of the moth had devoured the original occu- pants, but this could be considered only as an incidental occurrence. METHODS OF CONTROL. There is little doubt that in apple orchards which are sprayed with arsenicals for the codling moth and other common insect pests, many of the adults of this borer may be killed incidentally. The fact that beetles feed rather freely on exposed surfaces, especially on the wrinkled bark at the base of twigs where deposits of the poison from sprays collect and adhere, makes them susceptible to this means of control. The beetle has a rather prolonged feeding period before oviposition begins, and this affords a chance to kill it with poison sprays before it has provided for a succeeding genera- tion of borers. The borers, while small, may be found and removed from the trees very readily by paring away the bark over their burrows with a sharp knife. The burrows can be located without much trouble by the conspicuous castings which are thrown out, and also by the swollen, cankerlike appearance of the affected wood. (PI. IV, C.) Badly infested branches can often be removed without injury to the tree, and the borers within be destroyed by burning. Breeding places, such as are provided by neglected seedling apple trees, and thorn and wild crab apple thickets, should not be allowed to remain in the immediate vicinity of cultivated orchards. PUBLICATIONS OF UNITED STATES DEPARTMENT OF AGRICULTURE RELATING TO INSECTS INJURIOUS TO DECIDUOUS FRUITS. AVAILABLE FOR FREE DISTRIBUTION. Spraying Peaches for the Control of Brown Rot, Scab, and Curculio. (Farmers' Bulletin 440.) More Important Insect and Fimgous Enemies of the Fruit and Foliage of the Apple, (Farmers' Bulletin 492.) San Jose Scale and Its Control. (Farmers' Bulletin 650.) Apple-Tree Tent Caterpillar. (Farmers' Bulletin 662.) Round-headed Apple-tree Borer. (Farmers' Bulletin 675.) Rose-chafer: A Destructive Garden and Vineyard Pest. (Farmers' Bulletin 721.) Leaf Blister Mite of Pear and Apple. (Farmers' Bulletin 722.) Oyster-Shell Scale and Scurfy Scale. (Farmers' Bulletin 723.) Orchard Barkbeetles and Pinhole Borers, and How to Control Them. (Farmers' Bulletin 763.) Carbon Disiilphid as an Insecticide. (Farmers' Bulletin 799.) Aphids Injurious to Orchard Fruits, Currant, Gooseberry, and Grape. CFanners' Bulletin 804.) Important Pecan Insects and their Control. (Farmers' Bulletin 843.) Gipsy Moth and Brown-tail Moth and their Control. (Farmers' Bulletin 845.) Cranberry Insect Problems and Suggestions for Solving Them. (Farmers' Bulletin 860.) Information for Fruit Growers about Insecticides, Spraying Apparatus, and Important Insect Pests. (Farmers' Bulletin 908.) Control of the Codling Moth in the Pecos Valley in New Mexico. (Department Bulletin 88.) Walnut Aphides in California. (Department Bulletin 100.) Lesser Bud-moth. (Department Bulletin 113.) Life History and Habits of the Pear Thrips in California. (Department Bulletin 173.) Control of the Grape-berry Moth in the Erie-Chautauqua Grape Belt. (Department Bulletin 550.) Pecan Leaf Case-Bearer. (Department Bulletin 571.) Orchard Injury by the Hickory Tiger-Moth. (Department Bulletin 598.) Life History and Habits of the Mealy Plum Aphis. (Department Bulletin 774.) Apple Maggot or Railroad Worm. (Entomology Circular 101.) How to Control the Pear Thrips. (Entomology Circular 131.) Woolly Apple Aphis. (Office of the Secretary, Report 101.) FOR SALE BY THE SUPERINTENDENT OF DOCUMENTS, WASHINGTON, D. C. Important Insecticides. (Farmers' Bulletin 127.) Price, 5 cents. Insects Injurious in Cranberry Culture. (Farmers' Bulletin 178.) Price, 5 cents. Spraying for Apple Diseases and Codling Moth in the Ozarks. (Farmers' Bulletin 283.) Price, 5 cents. Insect and Fungous Enemies of the Grape East of the Rocky Mountains. (Farmers' Bulletin 284.) Price, 5 cents. Danger of General Spread of the Gipsy and Brown-tail Moths Through Imported Nursery Stock. (Farmers' Bulletui 453.) Price, 5 cents. Grape Leafhopper in Lake Erie Valley. 1914. (Department Bulletin 19.) Price 10 cents. (10) SPOTTED AFPLE-TREE BORER. 11 Studies of the Codling Moth in the Central Appalachian Region. 1915. (Department Bulletin 189.) Price, 10 cents. Homemade Lime-sulphur Concentrate. 1915. (Department Bulletin 197.) Price, 5 cents. Report on the Gipsy Moth Work in New England. 1915. (Department Bulletin 204.) Price, 30 cents. Food Plants of the Gipsy Moth in America. 1915. (Department Bulletin 250.) Price, 10 cents. Life History of the Codling Moth in Maine, 1915. (Department Bulletin 252.) Price, 10 cents. American Plum Borer. 1915. (Department Bulletin 261.) Price, 5 cents. Parandra Borer. 1915. (Department Bulletin 262.) Price, 5 cents. Cranberry Rootworm. 1915. (Department Bulletin 263.) Price, 5 cents. Dock False-worm: An Apple Pest. 1916. (Department Bulletin 265.) Price, 10 cents. Dispersicm of Gipsy Moth Larva' hy the Wind. 1915. (Department Bulletin 273.) Price, 15 cents. Miscellaneous Insecticide Investigations. 1915. (Department Bulletin 278.) Price, 10 cents. Terrapin Scale: An Important Enemy of Peach Orchards. 1916. (Department Bulletin 351.) Price, 15 cents. Cherry Leaf-beetle: A Periodically Important Enemy of Cherries. 1916. (Depart- ment Bulletin 352.) Price, 10 cents. Grape Leaf-folder. 1916. (Department Bulletin 419.) Price, 5 cents. Apple Leaf -sewer. 1916. (Department Bulletin 435.) Price, 5 cents. Pear Leaf -worm. 1916. (Department Bulletin 438.) Price, 5 cents. Solid-stream Spraying against the Gipsy Moth and the Brown-tail Moth in New England. 1917. (Department Bulletin 480.) Price, 5 cents. Mediterranean Fruit Fly in Hawaii. 1918. (Department Bulletin 536.) Price, 30 cents. Cranberry Girdler. 1917. (Department Bulletin 554.) Price, 10 cents. Striped Peach Worm. 1918. (Department Bulletin 599.) Price, 5 cents. Woolly Aphis of Apple. 1908. (Entomology Circular 20.) Price, 5 cents. Buffalo Tree-hopper. 1897. (Entomology Circular 23.) Price, 5 cents. Pear Slug. 1897. (Entomology Circular 26.) Price, 5 cents. Larger Apple-tree Borers. 1907. (Entomology Circular 32.) Price, 5 cents. Nut Weevils. 1908. (Entomology Circular 99.) Price, 5 cents. Two Destructive Texas Ants. 1912. (Entomology Circular 148.) Price, 5 cents. Mediterranean Fruit Fly. 1912. (Entomology Circular 160.) Price, 5 cents. San Jose or Chinese Scale. 1906. (Entomology Bulletin 62.) Price, 25 cents. Pecan Cigar Case-bearer. 1910. (Entomology Bulletin 64, Pt. X.) Price, 5 cents. Spring Cankerworm. 1907. (Entomology Bulletin 68, Pt. II.) Price, 5 cents. Lesser Peach Borer. 1907. (Entomology Bulletin 68, Pt. IV.) Price, 5 cents. Lesser Apple Worm. 1908. (Entomology Bulletin 68, Pt. V.) Price, 5 cents. Grape-leaf Skeletonizer. 1909. (Entomology Bulletin 68, Pt. VIII.) Price, 5 cents. Peach-tree Barkbeetle. 1909. (Entomology Bulletin 68, Pt. IX.) Price, 5 cents. Periodical Cicada. 1907. (Entomology Bulletin 71.) Price, 40 cents. Codling Moth in the Ozarks. 1909. (Entomology Bulletin 80, pt. I.) Price, 10 cents. Cigar Case-bearer. 1909. (Entomology Bulletin 80, pt. II.) Price, 10 cents. Additional Observations on the Lesser Apple Worm. 1909. (Entomology Bulletin 80, pt. III.) Price, 10 cents. On Nut-feeding Habits of Codling Moth. 1910. (Entomology Bulletin 80, pt. IV.) Price, 5 cents. 12 BULL,ET1N 886, U. S. DEPARTMENT AGRICULTURE. Life History of Codling Moth in Northwestern Pennsylvania. 1910. (Entomology Bulletin 80, pt. VI.) Price, 10 cents. Fumigation of Apples for San Jose Scale. 1909. (Entomology Bulletin 84.) Price, 20 cents. Grape Root-worm, with Special Reference to Investigations in Erie Grape Belt, 1907- 1909. (Entomology Bulletin 89.) Price, 20 cents. Papers on Deciduous Fruit Insects and Insecticides. 1912. (Entomology Bulletin 97, 7 pts.) Price, 25 cents. Life History of Codling Moth and Its Control on Pears in California. 1911. (Ento- mology Bulletin 97, pt. II.) Price, 10 cents. Vineyard Spraying Experiments Against Rose-chafer in Lake Erie Valley. 1911. (Entomology Bulletin 97, pt. III.) Price, 5 cents. California Peach Borer. 1911. (Entomology Bulletin 97, pt. IV.) Price, 10 cents. Notes on Peach and Plum Slug. 1911. (Entomology Bulletin 97, pt. V.) Price, 5 cents. Notes on Peach Bud Mite, an Enemy of Peach Nursery Stock. 1912. (Entomology Bulletin 97, pt. VI.) Price, 10 cents. Grape Scale. 1912. (Entomology Bulletin 97, pt. VII.) Price, 5 cents. Plum Curculio. 1912. (Entomology Bulletin 103.) Price, 50 cents. Life-history Studies on Codling Moth in Michigan. 1912. (Entomology Bulletin 115, pt. I.) Price, 15 cents. One-spray Method in Control of CodUng Moth and Plum Curculio. 1911. (Ento- mology Bulletin 115, pt. II.) Price, 5 cents. Life History of Codling Moth in Santa Clara Valley of California. 1913. (Ento- mology Bulletin 115, pt. III.) Price, 10 cents. Grape-berry Moth. 1912. (Entomology Bulletin 116, pt. II.) Price, 15 cents. Cherry Fruit Sawfly. 1913. (Entomology Bulletin 116, pt. III.) Price, 5 cents. Lime-sulphur as a Stomach Poison for Insects. 1913. (Entomology Bulletin 116, pt. IV.) Price, 5 cents. Fruit-tree Leaf-roller. 1913. (Entomology Bulletin 116, pt. V.) Price, 10 cents. W.VSniNCTON : COVKRN'MKNT PIUNTINO OFFICE ; 1920 UNITED STATES DEPARTMENT OF AGRICULTURE sdh^'^u % BULLETIN No. 887 Contribution from the Bureau of Entomology. •^J L. O. HOWARD, Chief. ^U^'^JL Washington, D. C. September 29, 1920 PEAR BORER.' By Fred E. Brooks, Entomologist, Deciduous- Fruit Insect Investigations. CONTENTS. Page. Introduction 1 History and distribution 2 Synonymy 2 Food plants 2 Nature of injury 2 The egg 4 Tlie larva 4 The pupa 5 The adult 6 Oviposition 6 Cannibalistic tendencies 7 Natural enemies 7 Methods of control 8 INTRODUCTION. The common name "pear borer," long ago bestowed upon the insect which forms the subject of this bulletin, is something of a misnomer, since the species attacks apple more extensively than pear. The name "apple crotch-borer" would be much more sug- gestive of the general habits of the insect. The species occurs over the eastern portion of the United States and in this region a few of the larvae usually may be found boring in the bark of almost any old apple tree that may be examined. So long as the attack is limited to only a few borers the effect on the general health of the tree is scarcely perceptible, but occasionally the insects concentrate in certain trees and there breed in numbers year after year. The injury which follows is cumulative, resulting in depleted vigor, and often in the death of a part or all of the tree. On the whole, this borer is more injurious than is commonly supposed. The following pages contain an account of this insect which is based very largely upon data collected by the Bureau of Entomology during the past five years. In gathering the data several badly infested apple orchards in Pennsylvania, West Virginia, and Missis- sippi were visited for the purpose of noting the extent of injury, con- ducting biological studies, and testing methods of control. Aegeria pyri Harris; order Lepidoptera, family Aegeriidae. 2 BULLETIN 887, U. S. DEPARTMENT OF AGRICULTURE. HISTORY AND DISTRIBUTION. The pear borer is a native American insect that was first described by Dr. Thaddeus W. Harris ^ in 1830. It is rather widely distributed in the eastern part of this country, having been recorded from the following States: Maine, New York, New Jersey, Pennsylvania, Michigan, Maryland, West Virginia, Virginia, North Carohna, South Carolina, Georgia, Mississippi, Alabama, Missouri, and Texas. The species no doubt occurs in several States not included in the fore- going list. SYNONYMY.^ Aegeria pyri Harris, New England Farmer, v. 9, 1830, p. 2; Amer. Joum. Arts & Sciences, v. 36, 1839, p. 313; Ins. Inj. Veget., 1841, p. 235; ibid, 2d ed., 1852, p. 256; ibid., 3d ed., 1862, p. 335; Harris's Corresp. (Scudder), 1869, p. 361; Walker, Cat. Lepid. Brit. Mus., Pt. VIII, 1856, p. 45; Packard, Guide Study Insects, 1869, p. 278 (and other editions); Thomas, 1st Rept. Nox. Ins. Illinois, 1876 (1878), p. 40; ibid, 2d Rept. 1877 (1878), p. 170; Stout, Rept. Kansas Hort. Soc. 1879 (1880), p. 88; Martin, Thomas's 5th Rept. Nox. Ins. 111. 1880 (1881), p. 107; Kellicott, Can. Ent., v. 13, 1881, p. 8; Grote, Check List N. Am. Moths, 1882, p. 12; Weed, Am. Nat., v. 23, 1889, p. 1108, fig.; U. S. Dept. Agi-., Div. Ent., Ins. Life, v. 4, 1891, p. 34; Saunders's Ins. Inj. to Fruit, 1883, p. 140, fig. 146; ibid., 2d ed., 1889, p. 140; Riley, Proc. Ent. Soc. Wash., V. 1, 1888, p. 85; Beutenmuller, Ann. N. Y. Acad. Sci., v. 5, 1890, p. 204. Trochilium pyri Fitch, 3d Rept. Nox Ins. N. Y., 1856, p. 349; Morris, Synop. Lepid. N. Am., 1862, p. 141. Sesia pyri Boisduval, Suites k Buffon, Nat. Hist. Lepid. Het., v. 1, 1874, p. 440; Smith, Cat. Ins. N. J., 1890, p. 289; Beutenmuller, Bui. Am. Mus. Nat. Hist., V. 8, 1896, p. 139; ibid., v. 9, 1897, p. 220; Beutenmiiller, Memoirs Am. Mus. Nat. Hist., V. 1, Pt. VI, 1901, p. 297. Aegeria koebelei, Henry Edwards, Papilio, v. 1, 1881, p. 196; Grote, New Check List N. Am. Moths, 1882, p. 12; Beutenmuller, Bui. Am. Mus. Nat. Hist., v. 4, 1892, p. 173. FOOD PLANTS. Larvae of the pear borer have been found commonly attacking pear and apple. Dr. J. B. Smith ^ records mountain ash (Sorhus ameri- cana) as a host plant. The writer has found the larvae attacking juneberry {Amelanchier canadensis) and thorn {Crataegus sp.), and in black knots caused by PlowrigTitia morhosa on wild and cultivated cherry (Prunus spp.). NATURE OF INJURY. Injury is done by the larvae feeding in the bark (PI. II, B), the bur- rows occasionally extending slightly into the sapwood. In infested orchards particular trees on which the moths deposit most of their eggs year after year usually occur. This preference for certain trees 3 Harris, T. W., Insects. In New England Farmer, v. 9. (1830-31), no. 1, p. 1-2, July, 1830. 3 The writer desires to express his thanks to Messrs. P. W. Mason and August Busck, of the Bureau of Entomology, for compiling and approving the synonymy as presented, and to Mr. Mason for supplying notes on the history, distribution, and food plants of the species under consideration. * Smith, J. B. Eeport of the Insects of New Jersey. In Ann. Rept. N. J. Sta. Museum for 1909, p. 519. 1910. PEAE BORER. 3 is probably due chiefly to the fact that the continuous working of generations of larvae in the bark causes a roughened surface (PI. II, A, C) which is attractive to the ovipositing adults. Injury may occur at almost any point above the ground, except on the smaller twigs. Larvae are commonly found feeding in the crotches and in places where there is a rough or broken surface caused by the previous feeding of their own kind or by other agencies. (PI. II, A, C; PI. III.) Favorite places of attack are around the borders of mechanical wounds in the bark, areas affected by sun scald and winter injury, and around the burrows of other species of borers. The larvae often develop within the excrescences of stem tumor {Bacterium tumefaciens) and black knot (PlowrigJitia morhosa), and about the borders of dead areas caused by pear blight (Bacillus amylovorus) . The writer has found pear borers attacking trees under the following conditions: In stem tumor of apple at Quincy, Pa., Adrian, W. Va., Demorest, Ga., Gadsden, Ala., and Love Station, Miss. ; in black knots on cultivated cherry and wild cherry (Prunus virginiana) at French Creek, W. Va. , and Winthrop, Me.; in wounds made by roundheaded apple-tree borers (Saperda Candida Fab.) in cultivated apple and juneberry trees (AmelancMer canadensis) at French Creek, W. Va., and Biltmore, N. C; in wounds made in cultivated apple and wild thorn trees (Crataegus sp.) by the spotted apple-tree borer (Saperda cretata Newm.) at French Creek, W. Va., and East Lansing, Mich.; in wounds made in apple trees by the flatheaded apple-tree borer (Chrysohothris femorata Fab.) at French Creek, W. Va.; in wounds made in apple bark by the yellow-bellied sapsucker (Sphyrapicus varius varius) at Moorefield, W. Va., and Demorest, Ga.; and in the edges of wounds made in the operation of grafting apple trees at Lancaster and Quincy, Pa., and French Creek, W. Va. Where the larvae were feedmg in the position last named they were interfering seriously with the union of stock and scion. Mr. E. B. Blakeslee, of the Bureau of Entomology, in notes furnished to the writer, states that at Winchester, Va., pear borers have been observed to complete the girdling of apple trunks partially encircled by collar blight. The damage done by a single borer is usually negligible, but the combined injury of a dozen or more borers may endanger the health or life of the tree. Often infested areas will occur at the upper part of the trunk where a number of branches originate (PI. II, A, C) and the branches will die one by one until the tree is ruined. Badly infested trees usually take on a scraggly, neglected appearance, the bark being rough and the growth slow. (PI. III.) Several orchards were visited during the present investigation in which it was not unusual to find from a dozen to a hundred borers working in a single tree. Under such conditions treatment of some kind for saving the trees is necessary. 4 BULLETIN 887, TJ. S. DEPARTMENT OF AGRICULTURE. THE EGG. The egg (PL I, A) is light glossy brown, oval, and flattened. One end is slightly truncate and one side distinctly concave, the depressed area bemg elongate and covering half the central surface of the side. Length 0.6 mm., width 0.3 mm. THE LARVA. The larva (PI. I, B, G; PI. II, B, D) is creamy white with a brown head, the average length of full-grown specimens being about 15 mm. and the width slightly more than 2 mm. The thoracic and ab- dommal segments from the first to the tenth are uniform in width. Segments 11, 12, and 13 taper abruptly to a blunt point. Through- out, the body is very sparsely clothed with short, stiff hairs. The larvae feed almost exclusively on the inner bark, although occasionally, where the bark is thin, they wiU gnaw slightly into the sapwood. When feeding in the excrescences of black knot or stem tumor they penetrate into all parts of the porous tissue; also, after feeding at the edge of dead areas on the tree, fuU-grown borers some- times penetrate into near-by decaying wood to construct their cocoons. Usually, however, feeding is confined to the bark of the trunk and larger branches. The completed burrows vary greatly in shape, but are usually in the form of broad, elongate, central spaces with short galleries leading off in different directions. Larval activity begins early in the spring and is marked by the ejection of fresh, reddish castings through the bark and, often, by a few drops of brownish water oozing from the wounds. Active feeding has been observed in West Virginia as early as the last of March. The larva winters in a silk-lined hibernaculum constructed in that part of the burrow where the larva chances to be when over- taken by the cool weather of autumn. (PI. I, G.) Some of the larvae attain fuU growth in the fall, and, after wintering in their hibernacula, construct cocoons in the spring without further feeding. These cocoons are f oimed by reshaping and adding to the hibernacula. In West Virginia there are both one-year and two-year larval periods, the duration of this stage of the insect's existence evidently depending somewhat upon food conditions but more upon the time in the season when the larva hatched. Apparently, larvae from early-laid eggs usually transform to adults the following season, having thus a one-year life cycle, while those from late-laid eggs live in the tree as larvae over two winters, having a two-year life cycle. On account of the difficulty of obtaining eggs, no individuals were reared under constant observation from eggs to adults. Over a hundred newly hatched larvae, however, were collected at various times in the summer and planted in apple trees where their develop- Bui. 887, U. S. Dept. of Agriculture. Plate I, Stages of the Pear Borer. ul. Eggs in natural position: much enlarged. -B, Larva, pupa, and cocoon; enlarged. C,D, Empty pupa cases projecting from the bark. E, Adult. F, Cocoons exposed by removing scales of apple barli; enlarged. G, Larva in hibernaculum; slightly enlarged. Bui. 887, U. S. Dept. of Agriculture. Plate II. Pear Borer and Nature of Injury to Trees. ^, Young apple tree showing characteristic injury at base of larger branches. B, Larva in burrow in apple bark; enlarged. C, Same as ^, showing injury more in detail. D, Borer working in excrescence on apple branch; natural size Bu!. 887, U. S. Dept. of Agriculture. Plate III. Trunk of Apple Tree, Showing Serious Injury by Pear Borer. PEAE BORER. 5 ment could be watched. Of these approximately 25 per cent had a one-year larval period and 75 per cent had a two-year larval period. It is possible that in the South, where the annual feeding season is longer, a one-year larval period may be the rule, while in the North, where the summers are shorter, there may be a constant two-year period for this stage. THE PUPA. Pupation takes place within an oblong-ovate cocoon formed of small particles of wood held together by a tough fiber of silk. The cocoon is always hidden beneath a scale of bark or wood fragments. (PI. I, B,/.) The pupa (PI. I, B), which is from 8 to 10 mm. in length, is at first 3"eIlowish white in color but soon changes to brown, the shade of color deepening as the imaginal stage is approached. The posterior margins of the abdominal segments are marked with rings of shghtly darker brown. The segments beyond the sixth are armed each with two rings of saw-tooth-like points which slant backward, the points on the front ring of each segment being strongest and extending at the ends shghtly beyond the spiracles. The points on segments beyond the eighth are stronger than those in front. When the adult is ready to issue the pupa works forward until only the tip of the abdomen remains in the cocoon and the posterior end projects out through the bark. The pupa case then splits across the head just back of the antennae, the rupture sometimes extending down the back of the thorax, and the moth issues suddenly and crawls away a short distance to harden. The empty pupa case is left projecting from the bark. (PI. I, C, D.) In one case observed the pupal stage covered a period of 23 days. Table I indicates the time when cocoons are constructed in various locahties and shows other vernal activities of the insect. Table I. — Spring activities of the pear borer. Year. 1912 1913 1915 1915 1915 1915 1915 1916 1916 1917 1917 1918 1918 Month and day. T-ocality. — • Remarks. June 5 Apr. 18 June 21 June 29 Apr. 25 Aj>r. 29 May 4 May 12 June 16 May 16 May 18 May 6 May 8 French Creek, W. Va.. Moorefleld, W. Va French Creek, W. Va.. do Cocoons plentiful. Moths Issuing. Cocoons found on old apple tree. About half of indi\iduals in cocoon. Of 60 individuals 4 had pupated. A few larvse have entered cocoons. Seveial cocoons found on apple tree. No cocoons. Larv;e plentiful. Less than half are in cocoons. No cocoons. Larvse plentiful. Cocoons with larva> in them. 1 pupated May IS, adult June 11. Cocoons present on apple trees. About half the larv?^ are in cocoons. Holly Springs, Miss . . . 'Biltmore, N. C French Creek, W . Va.. Winthrop, Me French Creek, W. Va. . do do do It may be seen from the foregoing table that at French Creek, W. Va., where most of the observations were made, the first cocoons were found on May 6 and the last on June 29. 6 BULLETIN 887, U. S. DEPARTMENT OF AGRICULTUEE. THE ADULT. The adult of the pear borer (PI. I, E) is a dainty little moth having an expanse of from 12 to 17 mm. and a length of from 7 to 10 mm. The wings are transparent, veined, bordered, fringed, and tipped with metallic purplish black or brownish black, the dark areas being partially covered beneath with yellow scales. The upper parts of the body are purplish black with white and yellow markings on the head, yellow markmgs on the thorax, and three more or less distinct yellow bands around the abdomen. The legs and underparts are heavily marked with golden yellow, the antennae and anal brush usually being marked with the same color. The colors throughout have a metallic luster, especially in fresh specimens. The scales rub off easily and there is considerable variation in color marks on this account, especially with specimens taken in the field. Table II, based on field and laboratory notes, indicates the time of year when moths are found issuing in different localities. Table II. — Time of issuing of pear borer moths in different localities. Year. Month and day. Locality. Remarks. 1911 1911 1912 1913 1915 1915 1915 1916 1916 1916 1917 1918 1918 1919 Aug. 15 Aug. 27 June 4 May 13 June 21 Apr. 29 Apr. 26 May 19 Julv 20 July 22 June 11 May 11 June 1 June 18 French Creek, W. Va . do do One moth issued. Do. First moth of season issued. Moth issued from cocoon found April 18. First moth of season issued. Moths were issuing from cocoons on apple. 2 moths had issued from cocoons. 8 moths issued from May 19 to June 2. 2 moths from material brought from Maine. Moth issued from cocoon from Michigan. Moth issued that had pupated 23 days before. First moth of season issued. Moths issuing in abundance. Many moths flying in apple orchard. Moorefield, W. Va French Creek, W. Va . Gadsden Ala Love Station, Miss French Creek, W. Va . .do ...do do .. . do do The foregoing table shows that the earliest record for moths is April 26, at Love Station, Miss.^ At French Creek, W. Va., where the rearing work was done, moths issued from May 11 to August 27, emergence covering a period of 108 days. OVIPOSITION. Many attempts were made to observe oviposition before the act was finally witnessed. Ovipositing females were first observed on June 18-19, 1919, in a large apple orchard at Qumcy, Pa. The weather at the time was clear and warm and oviposition took place during the brighter part of the day, being most active from 10 a. m. to 4 p. m. A series of badly mfested trees were located and by visitmg one tree after another a number of female moths were found hovering about the trunks and the bases of the larger branches. Their flight was rapid and wasplike but the light-colored markings of the ab- PEAK BORER. 7 domen and antennae made the insects fairly conspicuous against the dark background of bark. The moths had to be approached warily, for they were quick to take alarm at any careless motions made m their near presence. Wlien frightened they disappeared almost in- stantly and it was useless to try to follow them to another tree. In ovipositing the moths flew with a wavering, gliding mov^ement near to the tree with the antennie brushmg the bark. They alighted at frequent intervals and moved the tip of the abdomen back and forth as they crawled over the bark seeking for a rough surface or a crack in which to place an egg. Sometimes the entire abdomen would be inserted under a scale of bark, or into an opening, where it would be held motionless for a second while the egg was being laid. Evidently a female lays only one egg in a place at a time, but repeated visits of females to a suitable location result in the eggs being grouped to- gether by the end of the oviposition season. In one instance seven eggs were found in a heap at the bottom of a crack m the bark. The eggs are so small and inconspicuous that it is next to impossible to find them on the bark, even with an ordinary hand lens. Specimens of bark on which female moths had been seen to alight, but on which no eggs could be found in the field, were taken to the laboratory and placed under a microscope, where numerous eggs were easily dis- covered. CANNIBALISTIC TENDENCIES. In several instances a number of larvse of different sizes were col- lected in the field and placed with pieces of live apple bark m large vials to be taken to the insectary. Sometimes the larvae would be retamed in the vials for several days. In all such cases it was found that the larger larvae would kill and devour the smaller. That canni- balism is sometunes practiced under normal conditions is indicated by the fact that the larvae are always found occupymg burrows mde- pendent of one another, although eggs are frequently laid in groups. In a few cases small larvae were found devouring their own kmd on the trees. NATURAL ENEMIES. It is a common thuig to find burrows of the pear borer that have been opened and the occupants removed by woodpeckers, although the species of bird responsible has not been observed. The larvae and pupae are rather extensively attacked by parasites, perhaps 50 per cent of them being destroyed in this way. Table III gives a list of the hymenopterous parasites reared by the writer from the pear borer. In addition to the species named in the table, there is a record of another parasite, Stilhopoides smavora Roh.,^ reared from this host. "■' RoHWER, S. A. Descriptions of new parasitic Hymenoptera. In Proc. Ent. Soc., v. 15, p. 180- 188, 1913. 8 BULLETIN 887, U. S. DEPAETMENT OF AGRICULTURE. Table III. — Hymcnopterous parasites reared from the pear borer, Aegeria pyri Harris. Quaint- ance Series No. Name. Locality. Year. Determined hy — 9402 9407 9429 9431 9468 9499 9531 9539 Microbracon sp Phaeogcncs ater Cress Lissonota n. sp Itoplectis annulipes (Brulle). Macrocentrus n. sp Ephialtes acqualis ( Prov. ) . . . Tetrastichus sp Ephialtes acqualis (Prov.). . . Gadsien, Ala Love Station, Miss French Creek, W. Va. do do do do Quincy, Pa 1915 1915 1915 1915 1917 1918 1919 1919 R. A. Cushman. S. A. Rohwer. R. A. Cushman. Do. S. A. Rohwer. Do. A. B. Gahan. R. A. Cushman. METHODS OF CONTROL. The borers of this species work so near the surface that they can usually be removed with a sharp knile without difficulty and without much injury to the tree. It is not always easy to locate their bur- rows, but as a rule exuding frass and often a spot of moisture on the bark show where the insects are working. The rough places on the bark where groups of borers feed continuously can be pared away with a knife, the borers removed, and then the surface covered with coal-tar creosote tree paint or white-lead j^aint. Such a coat of paint is not only beneficial in protecting the wounds from air and moisture, but to some degree it prevents reinfestation by the insects. In one case in a badly infested apple orchard, in the State of Mississippi, several applications of a heavy paint to the rough areas where the borers were congregated served almost entirely to rid the trees, for a few years at least, of these insects. It was found also that aj^pUcation^ of viscous material, such as is used on sticky fly paper, made in tl summer to these rough places, entangled many of the moths whei they visited the places to lay their eggs. The shallow burrows of this species make it possible to kill manj of the borers by applications to the bark of penetrating oily or poison^ ous liquids. Kerosene emulsion and the standard emulsified oi sprays, with small quantities of sodimn arsenate added, when appliej to the bark over the burrows killed as high as 85 per cent of borers. Nicotme sulphate washes were less effective, but some of tl coal-tar products killed over 90 per cent of the borers without per- ceptible injury to the bark. The heavy oil materials, however, should always be used with caution until assurance is obtained of their non- injurious effects upon the trees. WA.SHINGTO.X : GOVERNMENT PRINTING OFFICE : 1920 UNITED STATES DEPARTMENT OF AGRICULTURE #■1 BULLETIN No. 888 i rMt>>C|ry Contribution from the Bureau of Entomology < ^S^^^» ^- O. HOWARD, Chief jTW^'^'We. Washington, D. C. PROFESSIONAL PAPER October 13, 1920 RESULTS OF EXPERIMENTS WITH MISCELLANE- OUS SUBSTANCES AGAINST CHICKEN LICE AND THE DOG FLEA. By W. S. Abbott, Entomologist, Enforcement Insecticide Act. CONTENTS. Page. Introduction 1 Chicken lice 2 Methods of testing 2 Oil preparations used as sprays 2 Oil preparations used as fumigants ; ; Oil preparations applied to roosts, drop- ping boards, and interior of chicken houses 4 Mercurial ointment 4 Chicken-lice powders 5 Miscellaneous preparations 5 Pyrethrum powder 5 Tobacco powders 6 Naphthalene 6 Miscellaneous powdered materials 6 Summary 7 Recommendations 8 Page. The dog flea 9 Methods of testing 9 Pyrethrum powder 9 Naphthalene 10 Chicken-lice powders 11 Miscellaneous powdered mixtures 11 Miscellaneous powdered materials 11 Tobacco powders 12 Sprays 12 Washes 13 Dips 13 Liquids used as fumigants 13 Summary 14 Recommendations 15 iM INTRODUCTION. connection with the enforcement of the insecticide act of 1910 a number of proprietary insecticides and the ingredients entering into their composition have been tested against chicken lice and the dog flea. These experiments have been briefly smnmarized and form, the basis of this bulletin. This work was done at the Insecticide Board's testing laboratory, located at Vienna, Va., which is under the supervision of Dr. A. L. Quaintance, of the Bureau of Entomology, and imder the direct charge of Mr. E. W. Scott.^ Assistance was also rendered at different times by Messrs. W. F. Turner, J. F. Zimmer, and W. B. Sill. Resigned July 15, 1919. 186609°— 20 2 BULLETIN 888, U. S. DEPARTMENT OF AGRICULTURE. CHICKEN LICE. Several species of lice were present on the fowls used in these tests, viz, the body louse, Menopon hiseriatum Piaget; the shaft louse, Menopon pallidum Nitzsch; and the large hen louse, Gonio- cotes ahdominalis Piaget. The most abundant species were the body louse and the shaft louse and these were always present in large numbers. METHODS OF TESTING. Three general methods of testing were employed: First, treating the individual fowls by dusting, spraying, dipping, or the "vent treatment" (see Mercurial ointment, p. 4-5); second, fumigation by confining the fowls in a box which had been painted or sprayed inside with the material to be tested; and, third, painting or spraying the roosts and dropping boards or the whole interior of the house. Dusting.— Two persons worked together in dusting, one holding the bird and spreading the feathers, while the other applied the powder with a shaker or small hand dust gun. Except where noted the pow- ders were well rubbed into the feathers. Spraying. — The liquids were applied with a hand sprayer and were also well rubbed in, unless otherwise noted. Dipping. — In the dipping tests the birds were held in the solution for from one-half to one minute and the head was submerged for about a second, once or twice. Verit treatment. — In these tests the materials were thoroughly rubbed into the skin of the fowl around, or just below, the vent. Fumigation. — In these tests the liquids, largely hydrocarbon oil mixtures, were applied to the bottom or bottom and sides of a box and the infested fowls were then placed in the box, which was cov- ered with two or three thicknesses of burlap. The birds were left in the box for from 5 minutes to 3 hours. Painting the house. — For these experiments the roosts and dropping boards, and often the whole interior of the chicken house, were painted or sprayed a short time before the birds went to roost, and the house remained tightly closed for that night. OIL PREPARATIONS USED AS SPRAYS. Tests with 22 different mixtures containing hydrocarbon oils, phenols, nitrobenzol, coal tar, wood-tar distillate, pyridine bases, and soaps showed that such preparations will kill practically all of the lice if lightly applied. The emulsions at dilutions not greater than 1 to 100 are also effective, but a dilution of 1 to 250 is of no value. Dipping tests indicated that these preparations are effective at about the same dilution. When dipped in or rubbed with the undiluted oils the fowls as well as the lice were killed. In two tests fowls dipped EXPERIMENTS AGAINST CHICKEN LICE AND DOG FLEA. 3 ill lime-sulphur concentrate (33° Baume), diluted 1 part to 14 parts of water, were completely freed from lice. The use of these mixtures can not be highly recommended for the following reasons: Too heavy an application may kill or seriously injure the treated birds; the plumage is left in a soiled and dirty condition and the oils retained on the feathers may have an injurious effect on the eggs in the case of sitting hens. Unless the dipping or spraying is done on a warm, clear day there is danger that the fowls may become chilled. OIL PREPARATIONS USED AS FUMIGANTS. Table I shows the results of tests with oU mixtures used as fumi- gants against chicken lice. Table I. — Results of tests with oil preparations used as fumigants against chicken lice. Two or more fowls confined in a sprayed or painted box far each test. Test No. Active ingredients.! Neutral coal-tar oil Coal-tar distillate, ammonia Phenols, coal-tar hydrocarbon oils. Coal-tar oils, phenols Coal-tar creosote oil Goal-tar distillate Phenols, light oil from coal tar Neutral tar oils, phenols, pyridine Creosote oil Kerosene oil, pine oil Coal-tar light oil Coal-tar oils, phenols Coal-tar oil, naphthalene, anthracene. Neutral coal-tar oil Coal-tar distillate Creosote oil, mineral oil Coal-tar distillate Creosote oil, anthracene Benzol, tar acids, naphthalene Coal-tar oils Coal-tar emulsion Treatment. Interior of box painted. do Bottom of box painted. do Interior of box painted. do .do. .do. .do. .do. Bottom of box painted. do do Interior of box painted. do do do do do do do.2 Exposure, Minutes. 5 10 10 15 30 30 Hours. 1 1 1 1 1 1 1 2 2 2 2 2 2 3 3 Per cent killed. 0 100 0 0 100 100 100 100 100 100 100 100 75 100 95-100 95-100 98-100 100 100 100 100 1 An active ingredient is one that is effective against chicken lice when used at a proper dilution. 2 Diluted 1 to 25. This table shows that almost all of these preparations were very effective against lice when the fumigation was continued for not less than 30 minutes. If_ only the bottom of the box was treated, an exposure of 10 or 15 minutes failed to kill any lice but when the whole interior of the box was painted or sprayed a fumigation of 10 minutes was effective. An exposure of 5 minutes under the latter conditions was of no value. Although effective, this method of treatment can not be recom- mended because of the injurious effect on the fowls. While no chickens were actually killed in these tests they were all more or less affected by the treatment as was shown by their very laborious 4 BULLETIN 888, U. S. DEPAETMENT OF AGRICULTURE. breathing and general weakness when released from the box. These effects passed away in 5 or 10 minutes, and in no case was any per- manent injury noted. This treatment also left the feathers some- what soiled, as the fowls in their struggles came in contact with the freshly treated surfaces of the box. In two tests where the bottom of the nest box used by a sitting hen was painted with this type of preparation it was noted that from 95 to 100 per cent of the lice on the hen were killed or repelled and the eggs apparently were not injured. While sufficient tests were not made to determine the practicability of this treatment for the control of lice on sitting hens, it appears to be worthy of further investigation. OIL PREPARATIONS APPLIED TO ROOSTS, DROPPING BOARDS, AND INTERIOR OF CHICKEN HOUSES. There have been many advocates of the theory that chicken lice can be killed by painting the roosts and dropping boards or the whole interior of the poultry house with various oil mixtures, the idea being that the vapors or gases arising from these paints would pene- trate the feathers of the roosting fowls and kill the lice. This method was given a very extensive trial and not one of the 42 differ- ent preparations tested was found to be of any value. These prepa- rations contained one or more of the following ingredients : Phenols, tar oils, hydrocarbon oils, creosote oil, carbon disulphid, wood-tar distillate, benzol, nitrobenzene, naphthalene, anthracene oil, and pyridine and were of the same general character as those discussed in Table I (p. 3). In these experiments the roosts and dropping boards or the whole interior of the house were thoroughly painted or sprayed just before the fowls went to roost and all doors, windows, and ventilators were closed during the first night. Five badly infested fowls were used in each test and at the end of one week examination was made for hving lice. Although an occasional dead louse was found on the dropping boards, in no case was the treatment of any practical value. Since many of these same preparations were found to be effective when used as fumigants in small boxes, it is apparent that the ineffectiveness of this house treatment is due to the fact that the fumes do not become concentrated enough to kill the lice. Tests were also made with lime-sulphur applied in the same way and this also was found to be of no value. MERCURIAL OINTMENT. Five proprietary and two homemade mercurial ointments were tested and found to kill from 95 to 100 per cent of the body lice present. These ointments contained from 3 to 26 per cent of mercury mixed with lard, vaseline, paraffin, or some semisoUd of this type EXPERIMENTS AGAINST CHICKEN LICE AND DOG FLEA. 5 and were applied by rubbing a piece about the size of a pea well into the skin around or below the vent of the infested fowls. This method of treatment has been thoroughly tested by G. H. Lamson and J. A. Manter.2 The active ingredient here is the mercury, since paraffin or vaseline applied in the same manner is of no value. CHICKEN-LICE POWDERS. Forty-five of the common proprietary powders generally sold as ''Lice Powder," "Lice Killer," "Lice Exterminator," etc., were tested and, with a few exceptions, were found to be effective if thoroughly applied. These powders are all of the same general type and contain one or more of the following active ingredients: Naph- thalene, nicotine, sulphur, pyrethrum, or phenols, and are mixed with lime, sand, talc, fuller's earth, or diatomaceous earth as a filler or carrier. Most of these active ingredients have been tested separately, and the minimum amounts that will l^e completely effective when very carefully and thoroughly applied by experienced persons are as follows: Naphthalene 10 per cent, nicotine 0.75 per cent, sulphur 20 per cent(?), pyrethrum 5 per cent, phenols 15 per cent (?). In some of these powders the actual percentage of one or even all of the active ingredients fell below the minima given above, but the combined action of all of the active ingredients was sufficient to make the preparation effective. MISCELLANEOUS PREPARATIONS. In two tests, in which powders containing nicotine, naphthalene, and sulphur were added to the dust baths furnished to lousy hens, 95 to 100 per cent of the lice were killed. This method of treatment is not, however, a very practical one, since all fowls do not dust themselves and the few that do not will eventually reinfest the whole flock. Nicotine, if used at a reasonable strength, was found to be effective as a dip or spray. Lime-sulphur added to the drinking water (1 teaspoonful to a gallon), and a powder composed of sulphur, salt, sodium carbonate, naphthalene, charcoal, and lime, which was given with the food, were found to be absolutely without any effect on the lice. PYRETHRUM POWDER.3 ^ In a series of 12 tests pyrethrum powder (insect powder, Dalma- tian insect powder, or Persian insect powder) was found.to be very effective, killing all of the lice within 24 hours. In dilution tests it was found that a powder containing 5 per cent of pyrethrum was effective if very thoroughly and carefully applied. ^Lamson, G. H., jr., and Manter, J. A. Some lice and mites of the hen. Storrs Agr. Exp. Sta.,StoiTS, Conn., Bui. 86. 1916. 8 Powdered flower heads of Pyrethrum cinerariaefolium and P. roseum. 6 BULLETIN 888, U. S. DEPARTMENT OF AGRICULTURE. Powdered pyrethrum stems were carefully tested and proved to be of no value against these insects. TOBACCO POWDERS. A series of 27 tobacco powders ranging from 0.025 to 5.26 per cent nicotine was thoroughly tested against chicken lice and the results of these tests may be summarized briefly as follows: From 0.75 to 1 per cent of nicotine is of some value against these lice, but 1.15 or 1.25 per cent is as low as can be relied on for the best results, and even then the powder must be very freely and carefully appUed. NAPHTHALENE. A large amount of work has been done with naphthalene in various forms as a remedy against chicken lice.^ The data gathered show that powders containing 20 per cent or more are very effective but can not be generally recommended, since as little naphthalene as 10 per cent may temporarily injure the dusted fowls and 60 per cent may kill them if the powder is well rubbed in. Finely powdered naphthalene was found to kill from 20 to 100 per cent of the lice when sifted over the backs of fowls after they had gone to roost at night. While in no case were the Hce completely eradicated by this treatment, and it is not as effective as the dusting of the individual birds, it is considered that this method may be of some value. I Numerous tests with naphthalene (6.49 to 100 percent) nest eggs, which have been frequently sold as remedies for Uce on laying and sitting hens, proved that they were of absolutely no value. When these eggs were placed in nests used by laying hens and allowed to remain from 1 to 3^ weeks, no effect could be noted on the lice and in several cases they were more abundant at the close of the experiment than at the beginning. They were useless against lice on sitting hens either when placed under the hen for two hours each week (as recommended), or when allowed to remain under the hen for the entire incubation period. Under the latter conditions, especially during hot weather, the hens were frequently driven from the nest and some of them were made very sick for two or three days. There was also considerable evidence to show that the eggs were seriously injured by the naphthalene, but not enough data are available to prove this absolutely. MISCELLANEOUS POWDERED SUBSTANCES. The following is a list of powdered substances that were found to be effective against chicken lice when used as dusts: Arsenic trioxid, barium fluorid, barium tetrasulphid,^ borax, boric acid, cloves, naph-