bt iy Mh i hay Pyabi ih , Pane hin abet: y i i wu Ni na Me a be t,o 7 eS 4 onal Muse ‘TWENTY-EIGHTH REPORT STATE ENTOMOLOGIST _ < ¥ CONTENTS PAGE Recent Illinois Work on the Corn Root-aphis and the Control of its Injuries... 1 IRE CeNbeALbICLES mi Ne bAe | eee None (check) ...:...sssee.i0. { 100 4 24 | #8 | 80 -[\ ea.) Bee SEIT, MOS hh sw nies cons REMY wr 200 38 56 66 68 3.69 Terpeneless lemon oil ........ 200 5 16 ao 27 2.19 MIBMUIBTERE O's Jit 5 ost Aug lp bow ots 200 59 68 69 71 4.75 15 The inference that terpeneless oil of lemon, produced by distilling off the more volatile terpenes from the ordinary oil, is more likely to be injurious to seed-corn than either the terpenes or the crude oil, is confirmed in an interesting manner by the results of experiments de- scribed by Dr. Whitten in the paper already referred to, in which he says: “At the present the trend of evidence tends to show that the [corn] grains bear immersion in the lighter oils without injury for much longer periods than in the heavier oils, and that the injurious after-effects of the latter are more pronounced than those of the Former.” As a general conclusion from this phase of my discussion, it may be said that my supposition that injury to planted seed-corn near Bloomington was owing to differences in the quality of the oils applied to the seed, was unfounded; and the cause of the observed variation in the effects of treatment must be sought elsewhere. As a further test of this conclusion we may take two sets of plant- ings, made at the insectary, in which the seed was treated with the same lemon oils, in mixtures of the same strengths, as were used by Mr. W.R. Scott, of Knoxville, and Mr. Frank Clark, of De Long, two Knox county farmers who reported serious injury to their seed. Sam- ples of the oils were obtained from these farmers with statements of the strengths used by them. Both bought their lemon oil in Knox- ville, but from different druggists. From the following table it will be seen that the insectary plantings of seed treated with these oils yielded results not materially different from those of the untreated check- plantings made at the same time; and the injury reported must hence: have been due to some local cause. Por PLANTINGS TREATED WITH ORDINARY LEMON OIL AND ALCOHOL AS USED By W. R. Scorr AND FRANK CLARK, EXPERIMENT OF SEPTEMBER 22, 1907 (2a aie ‘Per cent of plantsup | ees Insectary test 6& i) SO ae ao | September | October Ws 29 30 3 7 = Ants 583 861 455 695 Aphids 153 551 397 638 (Cea a ea Ants 564 819 877 635 Aphids 15 48 67 605 Cultivation ..... Ants 588 1,057 775 615 Aphids 159 529 378 335 From the above tabulation of the totals for these dates, it will be seen that while the oats plot contained about as many ants as any of the others, it contained only about a tenth of the number of root- lice found in the other plots at the first three counts; and these were no doubt migrants from the other plots or the descendants of such migrants. That it should have been so invaded by root-lice from ad- jacent plots by the 5th of July, over forty days after planting, that the infestation was now virtually equalized, is consistent with what we have found in similar cases, and it illustrates forcibly the fact that no field of corn is safe against this insect, even tho it may start free from the root-lice in spring, as long as adjacent fields are in- fested by it. Nevertheless, the larger yield of the oat-ground plot than that of the cultivation plot indicates, as we should anticipate, that rotation to oats makes in the end a more effective clearance of a field than a deep spring preparation of the soil. The fact must not be overlooked, however, that rotation to small grains acts slowly in spring, giving the ant-aphis inhabitants of the held sufficient food for some time, and ample opportunity to move, on foot and on the wing, to other fields, while deep and thoro stirring produces its effect at once, before the planting-time of the corn, killing and keeping down the food plants of the aphids, and so dispersing and disarranging the insect colonies as greatly to reduce the numbers of the root-lice. SUMMARY 1. The principal measures of protection against the corn root- aphis are rotation of crops; an early and deep plowing, followed by the repeated deep disking, of corn ground heavily infested by ants or known to have borne a crop injured by the root-aphis; and the use of repellent substances at planting-time, not by direct application to the seed, (which is dangerous to germination and early growth,) but by previous mixture with chemical fertilizers or other powdered sub- stances, to be dropped with the seed by means of a fertilizer-dropper attached to the corn-planter. Pages 1-4. 61 2. Articles already published in the Twenty-fourth and Twenty- fifth reports of the State Entomologist’s office show marked benefit in field experiments with deep plowing and repeated deep disking, and also as a consequence of the treatment of seed-corn with oil of lemon previous to planting, the last statement being based upon a single field experiment made in 1906. Pages 4-6. 3. List of operations described in this paper. Pages 6-7. 4. Experiments of 1907 show that wet weather at planting-time may either result in serious injury to the seed if repellents have been applied to it direct, or in such washing away of the repellent sub- stances that they produce no effect either on the seed or on the ants and aphids, the character of the effect apparently depending on the amount of rainfall and on its relation to the time of actual planting. Comparative experiments show that the injurious effects reported were not due, as at first surmised, to differences in the quality of the repellants used in different operations. Pages 7-21. 5. Laboratory experiments with a considerable variety of repel- lents applied by uniform methods to colonies of the corn-field ant in a special cage, showed that oil of tansy, oil of lemon, anise oil, tincture of asafetida, apterite, and vermicide were very strongly repellent, that kerosene, camphor, and coal-tar were less effective repellents, and that a considerable number of other substances tested were, if repellent at all, too slightly so to make them practically useful. Pages 21-41. 6. Additional field experiments made in 1908, in a spring season which proved to be very wet, resulted in no injury to the seed, and on the other hand in no benefit to the crop, flooding rains apparently washing away the repellents before they could take effect upon either the seed-corn or the insects. Pages 41-44. 7. Experiments made in 1910 with tincture of asafetida and oil of lemon, applied first to bone meal which was then dropped with the corn by means of a fertilizer-dropper attachment to the planter, and tested by the yield at corn-husking, showed a gain of 5.6 bushels per acre by the use of asafetida, and 10.8 bushels per acre by the use of oil of tansy, the first gain being obtained at a cost for materials and additional labor of thirty-four cents a bushel, and the second gain at twenty-seven cents a bushel. This result was the more encouraging since a very unfavorable spring caused an unusually poor stand and reduced greatly the general yield of corn. In a good corn season the gain would have been greater for the same cost. Pages 44-48. 8. Additional experiments with deep plowing and repeated disk- ing made in 1909 showed, in one case, a decrease, due to the treatment, of 43 per cent in the number of hills infested by ants, and 18 per cent 62 in the number of ants in the infested hills, and a decrease of 27 per cent in the number of hills infested by root-lice and of 9 per cent in the number of the root-lice themselves. In another case the number of hills infested by ants was reduced 71 per cent and the number of ants in the infested hills 83 per cent, the number of hills infested by root-lice 86 per cent and the number of root-lice in the infested hills 61 per cent. The same experiment showed that deep disking with a 20-inch disk was much more effective in diminishing the number of ants and root-lice than was the comparatively shallow disking of a 16-inch disk, the difference between the two methods of treatment being 34 per cent and 48 per cent in the number of hills infested by ants and by aphids respectively, and 13 per cent and 35 per cent in the number of these insects themselves. It was incidentally shown by this experiment that plowing to a depth of four inches does not suf- ficiently break up the nests of the ants, but that about 85 per cent of them may be broken up by plowing six inches deep, the remainder being at least broken into. Pages 48-53. 9. Observations made at night upon the movements of colonies of ants out of plots treated as above, and across furrows surrounding them, showed nearly two and a half times as many migrations from the plots deeply stirred as from the check plot. Migration lines across furrows plowed thru the center of each of the plots a week after planting, showed the amount of normal underground movements of the ants at this time. Making due allowance for this, it appears that the migration movement caused by the disturbance of the ants in treated plots was more than five times as great as this normal. Pages 53-55. 10. Plowing to a depth of six inches in a Galesburg field, in 1910, dispersed 55 per cent of the ant colonies in this field, and one disking after plowing dispersed 15 per cent more. Plowing six inches deep, disking three times, and rolling once, increased the yield of the plot nearly 25 per cent, at a cost of 22 cents a bushel. One 20-inch disk- ing followed by rolling gave all the advantages obtainable by addi- tional diskings. Pages 55-57. 11. Fall plowing and one spring disking are much more effective than spring plowing with no disking, the latter containing about three times as many ants and four times as many aphids as the former. Pages 57-58. 12. Change of corn ground to oats for one year, and fall-plowing of the oats stubble, gave a larger yield by 25 per cent than adjacent ground kept continuously in corn, this difference being accompanied by a root-louse infestation of young corn on the oats stubble about one-tenth that found in corn on old corn ground. Pages 58-60. OBSERVATIONS AND EXPERIMENTS ON THE SAN JOSE SCALE By STEPHEN A. FORBES, Stare Enromo.ocist The product of a considerable amount of field and laboratory work on the San Jose scale, much of which was done or begun under the immediate supervision of Jas. A. West, of my office staff, has lain for some time in the form of field notes and unfinished manu- script prepared by Mr. West before his lamented death in 1909. Lapse of time and the studies of others have made some of this material obso- lete; but such part of it as is here presented seems still useful and worthy of publication. The larger part of this paper relates, indeed, to a series of experiments not yet finished in 1909, but continued for two years thereafter, and reported now from the original field notes made partly by Mr. West, but mainly by W. P. Flint and L. M. Smith, who were his aids also in the earlier work. THe SAN JOSE SCALE ON FRUITS The San Jose scale infests not only the trunk, limbs, and leaves, but also the fruit of trees, often so marking the fruit as to make it unfit for sale. The danger of dissemination by this means is, how- ever, very slight, since the fruit itself or the parings from it are little likely to be so placed that young insects from them ean secure a lodgment or a suitable breeding-place. Our attention was especially ealled to this subject by a statement made in a bulletin of the Divi- sion of Zoology of the Pennsylvania State Department of Agriculture (Vol. IV, No. 7, November, 1906) which says, referring to the San Jose scale: ‘“‘Its abundance upon fruit of any kind need not be con- sidered alarming, as it can not possibly spread from fruit that has begun to ripen, because it dies on such fruits and can not reproduce from them. In Bulletin No. 8 of the same series, the same writer says: ‘‘This pest dies upon the fruit as soon as it ripens, and con- sequently there is no danger of disseminating or spreading it by this means.’’ October 3, 1906, Mr. West obtained three ripe Jonathan apples infested by the San Jose scale and kept them under observation on his office table. The scale multiplied readily on all of them. One of the apples was kept until November 29, at which time there were more than thrice as many well-developed San Jose scales on it as in the beginning, and most of them were alive. There were also many 64 newly set young upon it, and crawling young were seen just before it was peeled and eaten. December 11, 1907, two infested Ben Davis apples were taken from a package which had been in cold storage, and were placed in a warm room. Crawling young appeared on them after ten or twelve days, and some of these set and began to develop. One of these apples was kept two weeks and the other twenty-six days, and on the twenty- fifth day three crawling young were seen. Tue Lire History oF THE SAN JOSE SCALE IN ILLINOIS The only detailed experimental studies of the life history of the San Jose scale hitherto published are those of Theodore A. Pergande, in Howard and Marlatt’s comprehensive discussion of this insect, which appeared in 1896.* According to this report, four complete generations of the San Jose scale were regularly bred at Washing- ton, with the possibility of a partial fifth generation, and the astound- ing total of 3,216,080,400 was given as the possible number of de- scendants from a single female in a single season. Noticing that in these breeding experiments no account was taken of the difference in the number of annual generations, and conse- quently in the number of descendants produced by the first-born of the first-born of each generation, and by the last-born of the last- born of the same series—a consideration to which attention was ealled by the writer in 1906t—it was thought best to duplicate these studies at Urbana by methods which would bring this factor into view; and the problem was assigned to Mr. West for solution in 1908. Young and vigorous trees of the Ben Davis variety, on which the insect multiplies freely, were selected for the purpose; and glass rings 10 mm. in height and 35 mm. in interior diameter were fixed on the tree by means of paraffin or wax. Into one of these cylinders were put crawling young which were to start a new generation. The ring was covered either with a band of closely woven muslin or with glass. Cylinders with glass covers were not satisfactory, however, if so placed as to receive the direct rays of the summer sun, and the cells were almost invariably placed on a shaded part of the tree. It was our plan to rear two series of generations, starting in each case with the firsc and the last young of each brood of the insects. By taking the first-born of each brood, the generations were brought as close together as possible, giving the maximum number for the sea- son. By taking the last-born of each brood, the generations were separated as widely as possible, giving the minimum number for a season. *Bull. 3, N. 8., Div. Ent., U. S. Dept. Agr., p. 43. +The Corn Root-aphis and its Attendant Ant. Bull. 60, Bur. Ent., U. 8. Dept. Agr., 1906, p. 31. 65 Two hibernating females found producing young on May 30 (the first larve of the season to appear) were surrounded by eells as above described, and several of the first young born from them were carefully transferred on a camel’s hair brush to similar cells. In this manner two colonies were started from two similar parents; and they were carried thru the season, with a break, however, in one or them, caused by the melting in July of the paraffin which fastened the cell to a limb of the tree. The following diagram, devised by Mr. West, gives a complete view of the generations of the San Jose scale as they were thus reared in the open air at Urbana, IIl., during the season of 1908. The oblique lines indicate the growing period of the young insects; the horizontal lines, the period of maturity during which young were produced; and each bent, but unbroken, line represents the entire period of the generation, except that no account was taken of the life of the female after she had ceased to reproduce. An x on the diagram indicates the point of origin of a generation from the first-born young of a brood; and a 2, the point of origin of a generation from the last-born young. The W at the left indicates a hibernating female from which the series originated, and the W’s at the right indicate the groups of hibernating young produced. Fig. 1—Diagram of Annual Generations descending from one Hibernating Female. Urbana experiment. 66 One of the two hibernating females, born the preceding year, brought forth 109 young between May 30 and July 9. The first- born of these, belonging to the first generation of the year, produced 442 young from July 9 to August 22. The first-born of this second generation produced 491 young between August 14 and September 28, and the first-born of this third generation brought forth 233 young between September 17 and October 25. The earliest deseend- ants of this third generation were but partly grown when the winter came on. This fourth generation thus forms, of course, a part of the hibernating group corresponding to that with which the series started. There were thus four generations in a year of this series of the first-born, the number of young for each female of the suc- cessive generations being 109, 442, 491, and 235 respectively—an average of 319 to the generation. The last-born of this hibernating female appeared July 9, but these were lost by the accident mentioned above. Turning now to the second female of the hibernating generation, with which a second parallel series of breedings began, we find that this female produced between May 30 and July 15, 186 young of a first generation; that the first-born of this generation brought forth, between July 12 and August 24, 503 young of the second generation ; that the earliest born of these produced, between August 17 and September 29, 528 young of a third generation; and that from the first-born of these, 262 young of the fourth or hibernating generation were produced between September 19 and October 28. We thus have four annual generations of the first-born of this series also, with 186, 503, 528, and 262 young in the successive series—an average of 369 to the generation. The last-born of the hibernating female with which this second series began appeared July 15, and the last-born of the 498 young produced by this first-generation parent appeared October 1. These were not yet mature when the winter overtook them, and they form a part, of course, of the hibernating group. This gives us but two annual generations of the last-born series—but two ‘‘complete’’ gen- erations—and an average for the year of three generations when both first-born and last-born series are taken into account. By reference to the diagram it will be seen that there are three other lines of succession intermediate between these two extremes, and that each of these three lines represents three generations, mak- ing a total of sixteen generations for the five lines or an average, again, of three generations for the whole group. It will also be seen from this diagram that the generations represented by the San Jose scales present and alive in each month were as follows: May, the hibernating generation; June, the hibernating and first generations; July, the hibernating, first, and second; August, the first, second, and third; September, the first, second and third, and the fourth, 67 which is part of the hibernating generation to continue into the follow- ing year; and October, the second, third, and fourth. The hibernating assemblage contains representatives of at least three generations—the second, third, and fourth of the season—and these are of all ages, from young just fixed to those adult and quite ready to reproduce, or per- haps with a part or all of their brood already brought forth. A comparison of the intervals between birth and maturity shows, as might have been expected, shorter growth periods with the ad- vancement of the season, those of three successive generations of the descendants of the first overwintering female being 40, 37, and 34 days respectively, and those of the second overwintering female be- ing 438, 37, and 33 days. The reproductive periods of the successive generations differ little in length, but widely in productivity, the June and October generations containing fewer young than those brought forth in July, August, and September. The October young, indeed, were all produced during intervals of warm weather between periods of heavy frost. Evidently there can be no computation worth making of the actual or possible rate of multiplication of the San Jose scale which does not take account of the facts here given concerning the maximum and minimum numbers of the generations of the annual eyele, as shown by the first-born and the last-born series respectively, together with the proportion of each generation which are males and of the various rates of multiplication in different parts of the season. On the basis of the above data of West’s experiment, together with Per- gande’s percentages of males and females for the different genera- tions, P. A. Glenn, of my office staff, has worked out a day by day computation of births and deaths of both males and females for the period of 152 days during which reproduction was in progress at Urbana, with the result that there would have been produced by October 28, under ideal, optimum conditions, 32,791,472 descendants of a single female of the hibernating generation, of which 32,440,025 would have been still alive, 49.4 percent would have been a week old or less, 27.6 percent between one and two weeks old, 13.6 percent between two and three weeks, 5.5 percent between three and four weeks, and 1.11 percent more between four and five weeks old, and only 2 percent would have been mature. IT am assuming that, as the average growth period at Urbana was ot days, each week after birth would add about one fifth of the growth to adult size. The import of these figures may be better realized if they are expressed in the area which the total number of scales present in their various sizes October 28 would cover if placed in a single layer edge to edge. The average diameter of an equal number of male and female scales ig 1.1 mm. Taking this at 1 mm. and making propor- tionate allowance for the lesser size of the scales at the various stages of growth, we find that the entire product of reproduction alive Octo- 68 -ber 28 is equivalent to 8,803,283 adult scales—a number sufficient to cover 94.8 square feet. Even this estimate of a possible product of multiplication must be regarded as excessive if we take note of the fact, reported by Lowe and Parrott, of New York,* that an average of nearly 40 per- cent of the young scales may perish, apparently from mere physi- ological causes, without settling down to feed and secrete the pro- tecting scale. In Mr. West’s studies no attempt was made to trace the fate of any of the young except those chosen as parents of the generation to follow. These data of the Urbana experiment, obtained from a repro- duction season beginning May 30 and closing October 28, no doubt give us, however, a product very much smaller than might be worked out for southern Illinois, and probably a larger product than the northern Illinois normal rate of increase. More than 75 percent of the entire theoretical product of the season was but two weeks old or less October 28, and 8 percent of the scale insects alive at that date would have been but one day old. A relatively slight lengthen- ing of the period of reproduction would evidently have added enor- mously to the total number produced.t At Alton, crawling young were seen, in 1908, up to November 9, and in 1906 newly born young were found by R. D. Glasgow, in Williamson county, January 4. It is certain that young fruit-trees are much more rapidly destroyed in southern Illinois by San Jose scale than in the northern part of the state, and that the scale spreads more rapidly there and is more difficult to keep in check; and our experience also shows that in- festation by this insect is decidedly more injurious to susceptible plants in southern Illinois than it is in the central part of the state. TESTS OF ORCHARD SPRAYS Comparative tests of orchard sprays for the San Jose seale have been many times made, with the general result that the modern lime- sulphur preparations have come into very general use; and they seem likely to maintain this lead particularly by reason of a special value as fungicides which gives them an advantage over the petroleum preparations, their only real competitors. Nevertheless, careful ex- periments made by the office force of the State Entomologist from 1907 to 1911 seem still worthy of report, at least as a part of the permanent record. It was the principal purpose of these experiments, which, be- ginning in the fall of 1907, were planned to continue for five years upon the same orchards, to learn whether it was possible to redeem “Bull. 193, N. Y. State Agr. Exper. Sta., Dec., 1900, p. 355. , tIn another paper of this report, this fact is taken into account in a compu- See of the product of periods ten days shorter and ten days longer than the above, 69 orchards, already considerably or badly infested and situated in a generally infested territory, by measures which would make this economically worth while; and at the same time to show which of the various treatments used was the most successful and the most profitable. This purpose was in part defeated, however, by the failure of a company with which our contract for insecticide supplies was made to deliver the materials ordered in time for use in 1910. A breakdown in their factory of which we were not notified until it was too late to obtain supplies from another source, prevented our spraying these orchards in that year, and the San Jose scale conse- quently multiplied for a season without restraint. These orchards were saved notwithstanding, and are now in profitable use, but at a heavier loss and a greater expense than was necessary. EXPERIMENTS oF 1907-08 Two apple orchards, originally of 1200 and 1600 trees respect- ively, belonging to James M. Etherton and Homer Etherton, and ‘situated about eight miles south of Murphysboro, Jackson county, in southern Illinois, were selected for these tests. Both contained Ben Davis and Winesap apple-trees, so set in the spring of 1903 with two-year-old nursery stock that the orchards could be readily divided into similar experimental plots of sufficiently large size to permit the omission of the marginal rows of the plots and the use, for compara- tive purposes, of the central rows only. This, as will be shown, is an essential point in experiments of this description, since check and experimental plots placed side by side influence each other in a way to diminish the infestation of the margins of the former and to in- erease those of the latter. Infestation data were obtained by a critical inspection and grad- ing as described in an earlier article on ‘‘Comparative Exper iments with Various Insecticides for the San Jose Seale,’’ published in the Twenty-fourth Report of this office, pages 59-77. GRADING OF THE TREES The trees in the experimental orchards were carefully examined. one by one, November 11 and 12, 1907, by W. P. Flint and L. M. Smith, of the State Entomologist’s ie under the general super- vision of Mr. West. These inspectors worked together until it was evident that their procedure was uniform, after which they worked separately. The trees were graded on a scale of 10 degrees, the sev- eral grades having the following significance :— 1 signifies a tree infested by a trace of the San Jose seale—so very slightly infested that one must search to discover the insects: 2, a tree slightly infested—that is, having scattered scales upon sia nowhere clustered and yet fairly numerous; 3, a tree slightly in- 70 fested, with the insects showing a tendency in places to form clusters; 4, a tree with scales which show a decided tendency towards the for- mation of clusters, yet have parts but slightly infested; 5, a mod- erately infested tree, the scales being fairly abundant over the entire tree, and frequently forming groups or clusters; 6, a tree mostly moderately infested, but with some parts so badly infested as to be evidently suffering; 7, a tree considerable parts of which are badly infested, yet having parts infested to only a moderate degree; 8, a tree for the most part badly infested, but not incrusted, evidently suffering from the attack of the scale; 9, a tree badly infested, with some parts inerusted; and 10, an inecrusted tree. Cases where the scale was peculiarly distributed on a tree, making its classification. difficult, were decided by consultation. I am told, however, by the inspectors grading these trees, that, in practice, the grades between 1 and 10 were essentially estimates of the relative parts of the sur- face of the tree occupied by the scales, a grade of four, for example, indicating that twice as much surface was infested as in grade 2, and so on. It was found that the.trees could be readily graded, for the most part, on this scale and that the two men working inde- pendently would agree almost exactly in their estimate of the inten- sity of the infestation. The orchard containing originally 1200 trees was generally and badly infested, while the other, containing originally 1600 trees, was infested much more moderately. THE EXPERIMENTAL PLOTS Four insecticides were tested in this year’s operations in a way to compare for each the relative effects of a single fall application, a single spring application, and two applications, one in fall and the other in spring. The more heavily infested orchard (No. I), showing an average infestation of 7.4 degrees, was divided into five plots, one of which was reserved as a check, and the remaining four were treated fall and spring with the four insecticides. The check plot was six rows wide, plots 2 and 3 were fourteen rows wide, and plots 4 and 5, thirteen rows. Orchard No. II, the infestation of which averaged 3.9 degrees, was divided into nine plots, one of which was reserved as a check, the remainder being used to test the effect of a single application of: each of the four insecticides in fall and in spring. These plots were six rows wide, except the one nearest the check (7), which was made ten rows wide in order that a more liberal allowance might be made for the influence of the check plot upon the experimental plot adjoining than for the effects of adjacent experi- mental plots upon each other. INSECTICIDES The lime-sulphur mixture applied was the 15, 15, 50 preparation in general use by us in 1908. Fifteen pounds each of sulphur and i fresh stone-lime were boiled together for forty-five minutes, to make a strong solution, and this was then diluted with cold water to make fifty gallons. This ‘‘home-made’’ was compared with the ‘‘Rex’’ lime-sulphur solution of the Rex Company, of Nebraska, one part of which was used with eleven parts of water; with Sealecide, made by the B. G. Pratt Company, of New York, one part to twenty of water; and with the Target Brand Scale Destroyer made by the American Horticulture Distributing Co., of West Virginia, one part to twenty of water. These insecticides were applied with an Eclipse No. 6 pump carrying two leads of hose and 10-foot extension-rods. Each of the extension-rods was provided with a double Vermorel nozzle. The spray barrel was thoroly cleansed when the change was made from one insecticide to another. New nozzle caps having an aperture of about three-fourths mm. diameter were used, caps showing any wear being discarded. A constant strong pressure was maintained on the pump so that the spray was finely divided, ‘‘singing’’ as it left the nozzle. The various insecticides were applied in as thoro, uniform, —~ and fair a way as possible. In Orchard No. I the home-made lime-sulphur wash and the Rex lime-sulphur solution were applied to plots 2 and 3, November 19, 1907; the Target Brand and Scalecide on plots 4 and 5, November 23. The same plots were treated in the same way a second time March 11 and 12, 1908. Plots 7 to 10 inclusive, of Orchard No. II, were sprayed November 24, 1907, and plots 11 to 14 inclusive were sprayed March 12 and 13, 1908. FINAL GRADING October 5, 1908, all the trees in these orchards were graded a second time by Mr. Flint and Mr. Smith. Mr. Smith knew nothing of the details of the experiment. Altho Mr. Flint assisted in spraying the trees in November, 1907, he was not in the orchards again until the final grading of the trees, and the stakes which marked the plots had in the meantime been removed. Results of Treatment.—The results of a spraying operation may be stated in a way to compare the condition of the experimental plot at the time of spraying with its condition after the lapse of a season has made it possible to spray again. The ratio of these two stages of infestation, the original and the final stage, I have called the ratio of improvement or the ratio of loss or gain. If there is a check plot in the series, the rate of progress of its infestation during the season may be used to show the stage of infestation which the experimental plot would have reached if no spraying had been done, and the original infestation of this plot may be compared with this theoretical final stage, so ascertained. The ratio of these two stages I have 72 ealled the ratio of benefit.* In the following table both these ratios are shown for Orchard II, but the ratios of benefit are not given for ETHERTON ORCHARDS, 1907-08 OrcHARD I deca In- ; In- Loss |Perct. Porat festa- | festa- | (— f ; Plot exam: Treatment Date rae ce ( ae 2 - a f oe ined 1907 | 1908 | (+) | gain |Penetit 1 | 64 |None (check) 6.265 | 8.323 | —2.058|—32.9 2 | 148 |Lime-Sulphur Nov. 19 15 L., 15S, 50 W.| March 11/84 [5.8 |426 |431, 3 | 149 |Rex Lime-Sulphur, Nov. 19 5 1 part; water, 11 | March 11 | 8.8 5.95 | 12.85 |434.4 4 | 222 |Target Brand, Nov. 23 1 part; water, 20 | March 11 | 7.6 6.1 +1.5 |+20. 5 | 181 |Scalecide, Nov. 23 1 part; water, 20 | March 12 | 5.9 4.2 1.7 |-+-28.8 OrcHarD IT 6 86 |None (check) 3.44 | 4.56 |—1.12 |—82.5 7 | 109 |Lime-Sulphur 15 L., 15 S., 50 W.| Nov. 24 | 4.33 | 2.65 | 41.68 |438.4| 53.8 8 | 108 |Rex Lime-Sulphur, 1 part; water, 11 | Nov. 24 |404 | 2.65 | 11.39 |-40.4) 50.6 9 92 |Target Brand, 1 part; water, 20 | Nov. 24 | 5.37 | 3.66 |+1.71 |31. | 48.6 10 | 111 |Scalecide, 1 part; water, 20 | Nov. 24 | 4.88 | 3.02 | 41.86 |438. | 53.3 11 | 101 |Lime-Sulphur 15 L., 15 S., 50 W.| March 12 | 4.09 | 2.11 | 41.98 |+448.4) 61.1 12 | 104 |Rex Lime-Sulphur, 1 part; water, 11 | March 13 | 3.1 1.58 | 41.52 |+49. | 61.5 13 | 128 |Target Brand, 1 part; water, 20 | March 13 | 2.87 | 1.9 + .97 |-+-33.7| 50. 14 | 115 |Sealecide, 1 part; water, 20 | March 13 | 2.52 | 1.53 |+ .99 |+39.4!' 53.7 *The following is a convenient formula for use in this computation: b—a c-+e—d a —«x; a being the degree of original infestation of the check b—a c+ec a plot; b, the final infestation of that plot; c, the original infestation of the experimental plot; d, the final infestation of the same; and a, the ratio of benefit. 73 Orchard I, because the original infestation of this orchard was so high that an increase in plots 2, 3, and 4 at the rate of the check plot, would have completely incrusted these trees long before the end of the season. In other words, this orchard was too heavily infested in the beginning to be fit for full use in our experimental program. By an inspection of the foregoing table we find that two treat- ments, fall and spring, of the heavily infested orchard (the average infestation of whose experimental plots was 7.67 degrees) produced less effect than a single spraying, in either fall or spring, of the moderately infested orchard (the infestation of whose experimental plots averaged 3.9 degrees). The average improvement of plots 2 and 3 in Orchard I, the trees in which were twice sprayed with lime- sulphur preparations, was 32.7 percent, while that of plots 7 and 8 in Orchard II, once sprayed in fall, was 39.4 percent, and that of plots 11 and 12, once sprayed in spring, was 48.7 percent. This fact illustrates clearly the importance of early spraying before infestation becomes serious. In these experimental orchards there were but few crawling young up to August 1 on any of the treated trees which had originally been infested to 8 degrees or less; while on trees grading 9 and 10 degrees, crawling young were fairly abundant thru July, and began about the middle of August to cause a considerable re- infestation of surrounding sprayed trees, this dispersal increasing rapidly as the season progressed. While the insecticide treatment reduced a few of the completely incrusted trees to the 6th degree of infestation, many of those incrusted at the beginning of the experi- ment were partially or completely incrusted again by the end of the summer. The records of the checks in both orchards show a seasonal in- erease amounting to about one-third of the original infestation ; while the lime-sulphur treatments of plots 2 and 3 in November and March worked a reduction of about a third of the 8.5 degrees original in- festation of these plots. There was thus every reason to believe that a continuation of this program would save this orchard and virtually clear the trees of the scale; and this, I am informed, has since been done, both of these orchards being now productive and in good con- dition. The difference in effect between fall and spring spraying with lime-sulphur was shown by a comparison of the ratio of benefit (52.2 percent) of plots 7 and 8, sprayed in fall, with those of plots 11 and 12 (61.3 percent), sprayed in spring—a difference of about 17 per- cent in favor of the spring spraying. A similar comparison of the ratios of benefit of plots 10 and 14, sprayed with Scalecide, gives us a difference of less than 1 percent in favor of the spring spraying, and the ratios of plots 9 and 13, treated with Target Brand, show a differ- ence of less than 3 percent in favor of that insecticide. This is in accord with the general opinion that a spring spraying with lime- 74 sulphur prevents a fixation of the young scales more effectually than a treatment with the kerosene sprays. We also learn from the fore- going table that Scalecide was somewhat less effective than either form of lime-sulphur used; and that, of the two latter, Rex dip was about 5 percent more effective than the home-made solution. THE TRANSITION ZONE IN INSECTICIDE EXPERIMENTS In the Twenty-fourth Report of this office (1908) I called par- ticular attention to a generally neglected factor in field experimenta- tion with insecticides and other means of insect control;* namely, the fact that the effects of insecticide treatment of experimental plots are likely to be diminished and disguised by an invasion of these plots by insects migrating from adjacent untreated parts of the or- chard or field; and in that paper I illustrated this influence by a comparison of the degrees of infestation of check rows near experi- mental plots and experimental rows near check plots with the degrees of infestation of rows in each kind of plot at some distance from the other. The practical outcome of this comparison was a general rule that large enough plots should always be made in field experiments to give interior parts freedom from the influence of one plot upon another, and that the marginal parts of the plots should not be used in assembling data for comparison. This principle is amply illustrated in the results of the orchard experiments now under discussion. The check rows next the experi- mental plots were less infested by San Jose scale at the end of the season, and the experimental rows nearest the check plots were more infested, than were those at some distance from the line of division. This mutual influence of one plot upon the other varied somewhat with the degree of infestation, the neutral or transition zone of un- available trees being five rows wide in the less-infested orchard and six rows wide in that more infested. This transition zone was due, of course, to the movement of the young insects across the boundary between the plots, the marginal rows of the check losing more than they received from the experimental plot adjoining, and the experi- mental plot reversing the process. Even the outside rows of the check plot show, as might be expected, the same loss of infesting insects, so that only three rows were left available for use in a check plot six rows wide. The detailed data are given in the accompanying diagrams, one representing the less-infested orchard and the other the more-infested. The vertical lines on each diagram stand for the orchard rows in two adjacent plots, one the check and the other sprayed with home-made lime-sulphur wash. The heavy line between rows 6 and 7 is the *<< Spraying Apples for the Plum-eurculio,’’ pp. 94-99. — ra) dividing line between these two plots, and the heavy horizontal line marked with a zero stands for the average infestation of these plots before the spraying was done. The plus-figures at the left of each diagram indicate degrees of increase in the infestation of the check plot during the season following, and the minus-figures denote degrees of decrease of the infestation at the end of the season as a conse- quence of the treatment received. The broken line running diagonally across the diagram is drawn in a way to show, at the point where Check Flat (rows) L£xperimental [lot (rows) : FH Fic. 2.—Diagram showing Transition Zone, Orchard I, between Check Plot and Experimental Plot, Lime and Sulphur Treatment. Check Flot (rows ) Lxperunental -lot (rows) oss “Tt a ee eee eit ia = SoHE Ere e siege eees ola HEAT Fig. 3.—Diagram showing Transition Zone, Orchard II, between Check Plot and Experimental Plot, Lime and Sulphur "Treatment. 76 it crosses a vertical, the degree of loss or gain for that row, and the figures on the face of the diagram give those degrees precisely. It will be readily understood that if the check plot and the experimental plot were without influence one upon the other—if there had been no movement, that is to say, of the San Jose scale across the dividing line between the plots—then both these transverse lines on the dia- gram would have been virtually horizontal, their inner ends as widely separated in a vertical direction as are their outer ends, substantially as shown by the transverse dotted lines. The oblique part of each line marks the transition zone; and the upward turn of both lines at the left is an indication of the loss of insects from the outside rows of the check plots. EXPERIMENTS oF 1909 Continuing our operations in the Etherton orchards in 1909, we used for spraying experiments only Orchard II. Orchard I received a single spraying, for the mere control of the scale, March 18-23, 1909, with home-made lime-sulphur prepared as in the preceding year. According to the grading of October, 1908, the infestation of this orchard taken as a whole averaged 5 degrees on a scale of 10, and when graded again in January, 1910, its average was 2.39 de- grees, a gain in condition of 52.2 percent within the year as the resuit of the treatment. Orchard II was carefully graded March 11, 1909, and was di- vided into four plots. One was reserved as a check and the other three were sprayed, respectively, with home-made lime-sulphur, lime and sulphur manufactured by the Grasselli Company, and a petro- leum preparation known to the trade as San-U-Zay Seale Oil. The experimental plots were sprayed March 23-25 by Mr. L. M. Smith, assisted by one man and two boys, all entirely inexperienced in such work. Rains fell every day during the spraying, and showers and high winds on the 24th made the work ‘‘very unsatisfactory.’’ Prob- ably on this account, and because of the inexperience of the operators, the results of these experiments were much less favorable than those of the preceding and succeeding years. A second grading, for a comparison of conditions and an analysis of results, was made Janu- ary 17-18, 1910. From the following table it will be seen that the three preparations used were in practically the same class as to results produced, the home-made lime-sulphur, however, falling a little short of the other two. It was an interesting fact that the ‘miscible oil’’ preparation applied in March seemed practically equal in value to the solutions of lime and sulphur. wand dé ETHERTON ORCHARD EXPERIMENT, 1909 ORCHARD II No. No: : Average Per a Per et. of of Treatment Date (ie of loss of plot | trees March| Jan. |or gain Peo 1909 | 1910 fit 1 93 |None (check) 4.81 | 5.57 | —15.8 2 50 |Lime-Sulphur Grasselli Chem. Co. |March 25 5.56 | 4.93 | 411.3 | 23.4 3 190 |Lime-Sulphur home-made March 23-26 | 3.17 | 2.96 | + 6.6 | 19.3 4 94 |San-U-Zay Scale Oil |March 25 3.09 | 2.77 | 410.3 | 22.6 EXPERIMENTS OF 1911 After the omission of 1910, due to the failure of the insecticide company to fill our orders on time, both of the Etherton orchards were taken in hand in 1911 for experimental work. Orchard II was divided into four plots, which were sprayed, February 27 to March 6, with different brands of lime-sulphur. Orchard I was divided into five plots, four of which were sprayed March 6-9 with different lime- sulphur preparations, and the fifth with Scalecide. All the brands of manufactured insecticides were bought in the open market. The manufactured lime-sulphur solutions were all mixed with water at the rate of 1 gallon to 9. The home-made lime-sulphur contained 100 pounds of sulphur and 50 pounds of lime to 60 gallons of water, 9 gallons of this concentrate being diluted with cold water to make 50 gallons. Otherwise stated, the formula for this propor- tion was sulphur, 15, lime, 714, water, 50, differing from the home- made solution of the earlier experiments only in the smaller pro- portion of lime. The Scalecide was diluted in a ratio of 1 to 15. Owing to the condition of these orchards because of a lack of treatment the preceding year, it was not thought advisable or safe to leave any part without treatment, and no check plots were reserved. The work was in charge of Mr. Flint, assisted by Mr. Smith. The original grading was done by Flint and Smith, February 23, 1911, and the final grading, December 12 of the same year. In the absence of checks, only ‘‘ratios of improvement’’ could be com- puted. The home-made lime-sulphur, poured off after the settlement of suspended matter, gave an improvement of 40 percent, and the lime-sulphur of the Grasselli Company, 38.3 percent in Orchard I and 32.7 peréent in Orchard II, while the Rex lime-sulphur treat- ment resulted in an average improyement of 39.4 percent. 78 ETHERTON ORCHARD EXPERIMENT, 1911 ORCHARD I (AVERAGE ORIGINAL INFESTATION, 5.84) aie 5 cee x lan P nny No.of No.of batinent Deis verage intestationu| Per ct. of Beebe Feb. 1911/Dec. 1911] 84/2 it 183 |Lime-Sulphur Thomsen Chemical Co.!March 6 &7| 5.58 4.19 25. 2 185 |Lime-Sulphur Grasselli Chem. Co. |March 7 7.45 4.59 38.3 3 59 |Lime-Sulphur, home- made, with sludge |March 9 6.00 3.70 38.3 4 132 |Lime-Sulphur, home- made, without sludge|March 8 3.33 2.00 40. 5 144 |Scalecide March 8 6.86 5.04 26.5 ORCHARD II (AVERAGE ORIGINAL INFESTATION, 7.53) 6 142 |Lime-Sulphur Thomsen Chemical Co. |March 4 & 6} 5.73 2.57 26.7 7 151 |Lime-Sulphur Grasselli Chem. Co. |March 3 & 4} 8.48 5.71 32.7 8 140 |Lime-Sulphur Sherwin-Williams Co. |March 3 7.73 5.93 23.2 9 89 |Lime-Sulphur Feb. 27 & 8.20 4.97 39.4 Rex Co. Mar. 2 At the first glance, it would seem that suspended sediments in the home-made lime-sulphur were disadvantageous to it, the treat- ment of plot 4 with a clear solution giving a higher gain than that of plot 3 with a solution whose sediments were thoroly stirred up just before spraying. Since it so happened, however, that the origi- nal infestation of plot 4 (3.33 degrees) was only a little more than half that of plot 3 (6 degrees), it is altogether likely that this was the cause of the greater benefit to plot 4. Obvious differences may be noted in the results of the treatment with the several kinds of lime-sulphur tested, the Thomsen and the Sherwin-Williams’ preparations of 1911 being apparently in a differ- ent class from the Rex, the Grasselli, and the home-made preparations. Comparing the data of this table with those for 1907 and 1908, we find fairly similar results so far as the experiments themselves are fairly comparable. Scalecide, for example, gave us 28.8 percent of gain in Orchard I in 1907, and 26.5 percent in 1911. Rex lime- (i) sulphur shows 34.4 gain for Orchard I in 1907, on a plot the original infestation of which was 8.8 degrees, and a gain of 39.4 percent in Orchard II in 1911 vn a plot the original infestation of which was 8.2 degrees. It is true that the Rex plots in Orchard II agree less closely in the two years, the 39.4 percent of 1911 being matched by a gain of 49 percent in 1907; but a glance at the grades of infesta- tion explains the discrepancy, the 1907 plot having an average grade before treatment of 3.1 degrees, and the 1911 plot a grade of 8.20 degrees. We get a more obvious improvement, as a general rule, where the average infestation is the less. SUMMARY 1. Experiments with infested ripe apples show that the San Jose scale may live and reproduce freely on such fruits plucked from the tree and kept at ordinary room temperatures, and that living young may continue to be born under such conditions during a period of eight weeks. Infested apples taken from cold storage in December gave similar results, young being produced on these apples for twenty- five days. 2. Exact breeding experiments conducted at Urbana in 1906, in a way to distinguish thruout the season the descendants of the first- born from those of the last-born of each generation, gave two suc- cessive generations of the last-born series, in the complete year, and four such generations of the first-born series. A computation based on data thus obtained yielded a possible rate of multiplication under optimum conditions of 32,791,472 to 1 for the year. This total is only the 98th part of that of other investigators, who took no ac- count of diminished numbers of generations produced by late-born individuals. 3. Spraying operations with various preparations of lime and sulphur and with two brands of miscible oils justify the usual pref- erence for the sulphur solutions, especially because of their more prolonged effect when applied in spring. The home-made solutions were equally effective with those ready-made and requiring only dilution for use. These experiments also illustrate the great advan- tage of early spraying, before an orchard becomes heavily infested, and furnish evidence that spraying in spring is much more effective than spraying in fall, the ratios of benefit being some 20 percent greater. The possibility of redeeming and restoring a badly infested orchard and maintaining it in good condition with one or two spray- ings a year, was well established by these operations, LIFE HISTORY AND HABITS OF THE NORTHERN CORN ROOT-WORM (Diabrotica longicornis Say) STEPHEN A. FORBES, Strate ENTOMOLOGIST The life history of the common corn root-worm was established many years ago in its main features; and for a generation it has been commonly known that prompt rotation of crops is a complete preven- tive of the injuries of this insect to corn. Our knowledge of the subject is, however, still very deficient in detail, and the data of this paper, gathered from the miscellaneous observations and operations of several years, will help in some small measure to make good this defect. Breeding Experiments—Experiment 3054.