UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA CHEMICAL CONTROL OF THE GARDEN CENTIPEDE, SCUTIGERELLA IMMACULATA A. E. MICHELBACHER BULLETIN 548 DECEMBER, 1932 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA Digitized by the Internet Archive in 2012 with funding from University of California, Davis Libraries http://www.archive.org/details/chemicalcontrolo548mich CHEMICAL CONTROL OF THE GARDEN CENTIPEDE, SCUTIGERELLA IMMACULATA 1 A. E. MICHELBACHER2 Chemical treatment is often recommended for the control of soil pests, but the effectiveness of the chemicals is not always stated. This is par- ticularly true in the case of the garden centipede, Scutigerella immacu- lata (Newport) . With this in mind considerable work was done during 1931 to determine the effectiveness of various fumigants in the control of this pest. The materials used were carbon disulfide, carbon disulfide emulsion, paradichlorobenzene, calcium cyanide, naphthalene, potas- sium xanthate mixed with sulfur, hydrated lime mixed with sulfur, chloropicrin, carbon tetrachloride, and lead arsenate. PREVIOUS WORK Up to the present time most of the control work against the garden centipede has been done in greenhouses, although attempts at control- ling the pest under field conditions have been made. Methods used are soil fumigation, steam sterilization, poison baits, and installation of raised benches in greenhouses. The method most generally agreed upon as giving best results under greenhouse conditions is the installation of raised benches. McDaniel (8) 3 states that the cheapest and most satisfactory method of controlling the garden centipede consists in breaking the contact between the soil and subsoil. This, he further states, can most effectively be done by placing a matched board shield about 10 or 12 inches beneath the soil, or by installing benches with an air space between the bottom of the bench and the underlying soil. Compton (1) states that the "most effective permanent control is obtained by raising the benches and supplying clean soil." Filinger (4 ' 5) reports that where the installation of raised benches is practical it will solve the symphylid problem. Others who have recommended raised benches as an effective means of control are Herrick/ 6) Wallace et a£., (15 > and Riley. (12) i Received for publication September 13, 1932. 2 Research Assistant in Entomology. 3 Superscript numbers in parentheses refer to " Literature Cited' ' at the back of bulletin. 4 University of California — Experiment Station There is some evidence that the installation of raised benches is not always sufficient to prevent serious centipede damage. In Illinois' 7 * the pest is reported to have caused serious damage to chrysanthemums, roses, and tomatoes. However, this trouble was corrected by removing the soil, thoroughly cleaning the raised benches, and supplying new soil. Steam treatment (steam sterilization) 4 is another means often used in the control of the symphylid. According to McDaniel (8) "ordinary soil sterilization will kill the individuals near the surface, but will not affect those in the subsoil." Compton (1) found that steam treatment would kill all the symphylids in the upper part of the soil, but that in the case of ground beds, treatment must include the walks to be effective. Even then, he states, some of the symphylids and their eggs in the subsoil will survive, necessitating yearly treatments if they are to be kept under control. In his work in Ohio, Filinger (5) found that the buried-tile method of steam treatment was the only one that proved successful, and even that was effective only during the fall or winter when the centipedes are occupying the soil strata above the tiles. Further, the paths should be sterilized if possible. Thomas, (14) carrying on investi- gations in Pennsylvania, found steam treatment not wholly satisfactory even though the soil temperature rose above 160° F. He observed that many of the symphylids escaped by entering cracks in the concrete walls of the bed. Riley (12) in Indiana has observed that many centipedes are killed in the usual steam treatment as practiced by growers for the control of plant pests, such as lettuce drop and root nematodes, but that a sufficient number escape to maintain the infestation from year to year. Walton (16) working at Long Ashton, Bristol, England, reports attempts to control the centipede in infested areas by setting out tomato plants with roots and soil enclosed in grease-paper bags. The plants continued to grow until the paper decayed. The use of this method enabled the plants to gain size and hence greater resistance before being subjected to attack by the pest. There has been much conflicting evidence obtained where chemicals have been used in control of the garden centipede. McDaniel (8) reports that the use of chemicals in infested soil where crops are growing has not proved satisfactory to date. Compton (1) states that in Illinois a large number of chemicals have been tried, but none of the treatments so far devised is satisfactory under the conditions maintained in green- houses devoted to the production of sweet peas and other flowers. * Although the term "steam sterilization" is commonly applied to this method, complete eradication of life is seldom accomplished. Bul. 548] Control of the Garden Centipede 5 Considerable experimental work has been done with paradichloro- benzene. Feytaud, (3) carrying on work in France, reports his work with this material as not sufficiently conclusive to be recommended on a large scale. Wymore, (17) working in California, found paradichloro- benzene in several cases to be very effective, but under field conditions it did not prove so promising. This may have been due to the fact that the soil was not packed sufficiently at the time the material was applied. Also, the soil temperature may have been too low. According to the investigations carried on by Filinger (5) in Ohio, this material gave good results if applied in the summer or early fall when the symphylids are still in the subsoil. The surface soil is first removed and the paradi- chlorobenzene scattered over the subsoil at the rate of 1 pound to 100 square feet. After this the surface soil is again replaced. Essig, (2) working in California, reports that paradichlorobenzene proved unsatis- factory on the whole, probably because of the very porous condition of the peaty soils in which the experiments were made. In Oregon (10) paradichlorobenzene is reported as having a repellent effect and as killing a few of the symphylids. In Pennsylvania (14) this material is said to be ineffective in killing the centipede but has a repellent effect upon it. Riley, (12) carrying on investigations in Indiana, used paradi- chlorobenzene as a barrier between the surface soil and the subsoil. The work was done in the fall when the centipedes were in the subsoil and for a few weeks after treatment none could be found in the surface soil which was planted to lettuce. However, the centipedes soon began to work through the barrier, and within a year 's time large numbers could be found. Filinger (5) has used calcium cyanide against the garden centipede with success. He states that the material used at the rate of 1 pound to 75 square feet is effective when scattered over the subsoil and covered quickly. Wymore, (18) working with the same material under field con- ditions in California, found it to give only temporary relief. In Oregon 0) calcium cyanide is reported to have given practically no control. Crude naphthalene has been used but has not proved to be of any value. Work in Pennsylvania (14) has shown naphthalene flakes to act as a repellent. Herrick (G) has used carbon disulfide emulsion effectively against the centipede. A solution of miscible carbon disulfide was used at the rate of 1 gallon to 100 gallons of water for drenching the soil that had just been brought in from the fields and placed in the beds. He states that the dilute mixture used at the rate of 100 gallons for a bed 3 feet wide 6 University of California — Experiment Station and 160 feet long appeared to result in 100-per-cent mortality. Exami- nation of the soil for centipedes was made about 12 hours after the bed had been treated, and Herrick remarks that probably some of the centi- pedes escaped, although the examination was thorough. Filinger (5) carried on some work with carbon disulfide emulsion, and his investiga- tions showed that when it was used as a surface application it failed to control the pest. His stock mixture was prepared by dissolving 5 pounds of fish-oil soap in 3 gallons of water, to which 6 gallons of carbon disul- fide were added, and the mixture emulsified. The heaviest application consisted of 3 quarts of the stock emulsion to 50 gallons of water, applying the dilute mixture at the rate of ^ gallon to each square foot of surface. After such treatment the plants grew vigorously for a while, but before the crop was mature centipedes came from the subsoil and destroyed the plants. Work in Oregon with carbon disulfide emulsion showed it to act as a repellent for a short time. Riley (12) used carbon disulfide emulsion against the centipede but reports it as being ineffec- tive. Filinger's (5) investigations in Ohio showed that carbon disulfide proved satisfactory in the control of the symphylid when used at the rate of 1 ounce to a square foot. He applied it by sprinkling over the subsoil at the time of the so-called ''making up" 5 process. He also reports the material as being very effective when ''drilled" on both sides of the plant rows at the rate of 4 pounds to 150 feet of row. For good results it should be drilled in about 4 inches from the base of the plants and about 2% inches below the soil surface. Riley, (12) investi- gating in Indiana, has reported hand injections of carbon disulfide at the rate of 1 ounce to a square foot as giving desirable results. He (11 ' 12) also obtained good control when carbon disulfide was volatilized with steam and passed under pressure into ground benches which had previ- ously been equipped with tiles for steam treatment. Sheaffer (13) gives the same recommendations as Riley. It has been reported from Indiana (15) that creosote acts as a repellent to the centipede in ground benches, either applied as a light spray when spading or as a dressing, using sand soaked with creosote and turning the sand under. In Oregon (9 ' 10) it was found that poisoned corn (corn soaked for 48 hours in bichloride of mercury, 1 pound to 600 gallons of water) , caused the death of some centipedes, and that for several weeks after planting there were fewer centipedes in a treated area than in the check plot. 5 In the "making-up" process the surface soil is removed to a depth of 6 to 8 inches, then a layer of manure is placed on the subsoil, and the surface soil is re- placed on top of the manure. Bul. 548] Control of the Garden Centipede 7 It is also reported that most of the corn was eaten by the symphylids. In Pennsylvania (14) a poison bait consisting of bran and syrnp mash containing sodium arsenite proved nonattractive to the centipedes. It is also reported that slices of potato covered with paris green were non- attractive. Wymore/ 17 ' 18) in California, has reported flooding as being very effective in the control of the garden centipede. His work in asparagus fields indicates that flooding to the depth of a foot or more for two or three weeks, during the winter when the plants are in a dormant condi- tion, gives good results. Many investigators are of the opinion that frequent cultivation is unfavorable to the centipede and that great care should be used to avoid transporting the organism into greenhouses. Sanitary practices and sterilization of soil before placing it in beds should be carried out where possible. CONDITIONS UNDER WHICH EXPERIMENTAL WORK WAS CONDUCTED A field infested with centipedes at Clarksburg, California, was divided into 39 plots by ditches which were approximately 30 inches deep. These ditches were partly filled with water, which served to isolate the plots. The water was supplied by a pump, and the height to which it might rise was regulated by an overflow drain at the lower end of the experimental area. The plots varied in size from 12 x 12 feet to 18 x 18 feet. Five plots so selected as to represent an average infesta- tion were set aside as checks. Figure 1 shows the arrangement of the plots and the location of the checks. There was some cracking and Assuring of the banks of the plots, and in order to compensate for this an extra heavy application of material was placed along the edge. In the case of the solid materials an amount equal to 10 grams per linear foot of perimeter was distributed in a furrow as close to the edges as possible. In the case of materials which were injected into the soil a double dose was given around the edge of the plots treated and in the case of carbon disulfide emulsion the banks were saturated. Cracking and Assuring of the banks permits the centi- pede, as well as other soil-infesting organisms, to escape treatment in spite of the fact that an extra amount of the fumigant is applied to the edge of the plot. To avoid this the edges of the banks should be well trimmed of all loose soil. The soil in the experimental area is developed from a mineral sedi- ment deposited by the Sacramento River. The surface soil to a depth 8 University of California — Experiment Station of 8 to 10 inches is a brown silty clay, finely granular in structure, with a high content of organic matter. The subsoil is a light-gray silty clay with rust-brown and dark-brown stains and is permeated with countless fine cavities left by decaying roots, which serve admirably as runways for the garden centipede. At the time the chemicals were applied examinations showed that most of the centipedes were concentrated in the surface foot of soil, the majority being found in the first 9 inches. Molted skins were found at a / 1Z (Check) 13 21 25 51 (Check) 35 I 11 1 M 23 26 33 36 3 10 15 ZZ (Check) 27 3Z 37 4 9 16 Z1 Zd 31 36 J d (Check) 17 ZO Z9 30 (Chech) 39 6 7 18 19 1 inch = approximate ly 26 feet Fig. 1. — Arrangement of the experimental plots. depth of over 2 feet, but they are very resistant to decomposition and had doubtless been shed in the fall before the plots were treated. At that season of the year the usual practice is to lower the water table by draining the ditches throughout the district, and as the soil starts to dry out the centipedes undoubtedly tend to move deeper into it. At the same time the plots were treated the water table, whose position varies during the year from 2 to 5 feet down, was within 2 to 3 feet of the ground surface. The surface soil at this time contained an optimum moisture content for plant growth. Since the effectiveness of some of the chemicals used depends upon the temperature of the soil, records of soil temperature were taken for a period of approximately 2 months before and after treatment. Bul. 548] Control of the Garden Centipede TABLE 1 Maximum and Minimum Temperatures for 3 and 10-inch Soil Depths on Days That Temperature Eecords Were Taken Date (1931) Temperature readings 3 inches deep Temperature readings 10 inches deep Maximum Minimum Maximum Minimum °F » F °F °F 96.0 69.0 75.5 73 5 96 70 5 77.5 75 5 92 69 76.5 75.0 94 71 5 78.0 76.0 94 71 80 77.5 86.0 67.5 79.0 76.5 85.0 67.0 77 75 930 700 79.0 76.5 96.0 74 82.0 79 5 98.0 76 82.5 80.5 82 68 5 79.5 77.0 88 690 79.0 76.5 93 69 79 77.0 81 63 76 64.0 81.0 62.5 77.5 74.5 83.5 665 76 74.0 85.5 68.0 77.0 75.0 94 5 69.5 78 76.0 96 5 72 82.5 79.5 97.5 72 82.0 78.5 82 5 67 75 5 72.0 80 68.5 74 72.0 81.0 63 72 5 70.0 82 5 63 73 70 74 62.5 70.0 68.0 78.5 64 71 69.0 79.0 60.0 70 67 73.5 59.0 69.0 66.0 73 59 68.0 66 72.5 59.5 67.5 65 5 73.5 60.5 68.0 65.5 77 60.0 68.0 65.5 66.0 59 5 64.5 63 72.0 53.0 63.0 60.5 71.5 53.0 63.0 60 5 55.0 43.0 55.0 52 50 5 47.0 49.0 485 June 24 June 25 June 30 July 1 July 9 July 14 July 15 July 16 July 21 July 23 July 28 July 29 July 30 August 6 August 13 August 14 August 20 August 21 August 25 August 26 September 2... September 3.. September 10 September 11 September 15.. September 16.. September 24 . September 29.. October 7 October 8 October 9 October 16 October 21 October 30 November 4... November 18.. December 2 ... In the center of the experimental area a Bristol's recording ther- mometer was installed. Soil temperatures were recorded for 3 and 10-inch soil levels. Table 1 shows the maximum and minimum tempera- tures for both levels on days that records were taken. The maximum temperatures were usually reached rather late in the afternoon (4 to 6 p.m.), while the minimum temperatures usually occurred between 5 and 8 a. m. The maximum and minimum temperatures for the 10-inch soil level lagged behind that of the 3-inch level. 10 University of California — Experiment Station METHODS OF APPLICATION In treating the plots four methods of applying materials were used, namely, in furrows, broadcast and spaded under, injected or poured into holes in the soil, and mixed with water and poured over the soil. Application in Furrows. — The nonliquid soil fumigants with the exception of naphthalene, lead arsenate, and the lime and sulfur mix- ture, were distributed in furrows 18 inches apart from center to center and approximately 6 inches deep. The furrows were made with a narrow shovel and were approximately 10 inches wide at the bottom. In applying the fumigants the material was scattered over this entire area. To insure even distribution of the chemicals the total amount necessary for any plot was first weighed out, then weighed into equal parts corresponding to the number of furrows for each plot. After the application of the fumigant the furrows were partly filled and the soil packed by walking back and forth on it. After this the furrows were completely filled and further packed. The chemicals applied in the above manner were paradichlorobenzene, calcium cyanide, and a mixture of equal parts of potassium xanthate and sulfur. Paradichlorobenzene and calcium cyanide were applied at the rate of 300, 600, and 900 pounds to the acre, while the potassium- xanthate sulfur mixture was applied at the rate of 1,200 and 2,400 pounds per acre. All treatments were run in duplicate. The effective- ness of these chemicals is shown in table 2. Broadcasting. — Chemicals applied by broadcasting were naphtha- lene, a mixture of equal parts of lime and sulfur, and lead arsenate. The exact amount of material needed for each of the plots was first weighed out, then to insure an even broadcast of the chemical over a plot, the material was again weighed into a number of equal smaller amounts. This greatly facilitated the even distribution of the materials over the plots. As soon as the chemical was broadcast it was spaded under. Naphthalene was applied at the rate of 1,200 pounds to an acre, lime sulfur mixture at the rate of 1,200 pounds per acre, and the lead arsenate at the rate of 300 pounds per acre. These treatments were also run in duplicate. Their effectiveness likewise is shown in table 2. Injection. — Chemicals injected or poured into holes in the soil were carbon disulfide, carbon tetrachloride, and chloropicrin. The first two were poured into holes, while the last was injected into the soil with a mechanical soil injector. The apparatus used for applying the carbon disulfide and carbon tetrachloride consisted of a prod for making the Bul. 548] Control of the Garden Centipede 11 holes, and a funnel soldered to a length of pipe which could be inserted into the holes. The funnel was equipped with a small container which had a capacity of slightly more than 2 ounces. This small measuring device was suspended in the center of the funnel by means of a small cross rod so that it could be inverted to spill the measured liquid into the funnel, whence it passed down the tube into the soil. A dosage of 1.92 fluid ounces filled the container to the level of the supporting cross rod. A large teapot was used in filling the measuring device. In the case of the 0.96-ounce treatments the chemical was measured into a graduate and then poured directly into the funnel, whence it passed on into the soil. After the f umigants were applied the holes were filled and thoroughly tamped. Carbon disulfide was applied at the rate of 1.92 fluid ounces per hole in 6-inch and 9-inch holes, and at 0.96 fluid ounce per hole in 6-inch holes ; carbon tetrachloride was applied at the rate of 1.92 fluid ounces per hole in 9-inch holes, and chloropicrin at the rate of 0.13 fluid ounce (6.55 grams) per injection in holes 6 to 9 inches deep. (See table 2) . All treatments were made in duplicate. The distance the gas would have to spread through the soil could have been minimized by staggering the holes. This was not done because it was thought best to study the results at the maximum distance encoun- tered with an 18-inch spacing. Table 2 gives the effectiveness of these materials. Surface Application of Emulsion. — Carbon disulfide emulsion was the only material mixed with water. The stock emulsion used contained 65 per cent of carbon disulfide. 6 The stock emulsion was diluted 1 to 300, and 5 gallons of this dilute mixture per square yard was applied with a sprinkling can. Two plots were treated with this material, and in order to secure an even application of the emulsion the plots were staked out in yard squares. After treating this area the edges of the plots were well wetted with the emulsion. COMPARATIVE EFFECTIVENESS OF THE CHEMICALS USED A population study of the plots was made before treating them to determine the effectiveness of the various materials used. Two cylin- drical samples 6 inches in diameter and 9 inches deep were taken from each plot, and a complete count of the centipedes in each sample was made. The figure used for initial population was the average of the 6 This emulsion is a product of Wheeler, Keynolds, and Stauffer Company, San Francisco. 12 University of California — Experiment Station count in the two samples. This gave a rather complete picture of the distribution of the centipedes through the experimental area. From time to time after the treatment of the plots, examinations were made for centipedes. These showed that carbon disulfide, carbon disul- fide emulsion, carbon tetrachloride, chloropicrin, and calcium cyanide killed rapidly. At least in the case of the first three mentioned, almost all killing appeared to take place in the first 24 hours. The lethal effects of naphthalene and paradichlorobenzene were drawn out for a rather long period. The killing action of paradichlorobenzene lasted until ap- parently all the centipedes were killed, but the effect of naphthalene wore off before the centipedes were eradicated. Approximately 2% months after treating the plots a second popula- tion study was made. This should have allowed enough time for all the materials used to have exerted their full killing power. When the final count was made the population was so low on some of the treated plots that it seemed advisable to take larger samples than had been used in the first counts. Accordingly, 10 shovelfuls of soil were examined from each plot. The formula used in calculating the effectiveness of the treat- ment was designed to correct for the difference in volume between the samples of the first and second counts. Since all of the chemicals except carbon disulfide emulsion were ap- plied under favorable conditions, they could not be considered effective unless they resulted in nearly a 100 per cent kill. The conditions neces- sary for good results with carbon disulfide emulsion are discussed on page 15. Table 2 shows the effectiveness of the various materials used. The effectiveness of the chemicals is expressed as a percentage of the initial population killed ; the following formula 7 was used in calculating these results : T-rt, , . (initial population - corrected final population) X 100 Effectiveness = — . .,. . =— r? — — initial population The effectiveness of the individual treatment, and also the average of duplicate treatments are shown in the table. Although most of the plots receiving the same applications check rather closely with one another, several exceptions to this may be noted. In plot 18 (naphthalene 1,200 7 The corrected final population of each plot was found by multiplying the actual final population by a factor which included two corrections: (1) the average ratio of initial and final populations of the five check plots, and (2) the ratio of the volume of the earth used in the initial and final population counts of the test plots. In this way the actual increase in population during the long interval (2% months) between counts and the unavoidable change in method of counting were both allowed for, so that all counts are reduced to the same basis. Bul. 548] Control of the Garden Centipede 13 TABLE 2 Effectiveness of Various Chemical Treatments in Killing the Garden Centipede Chemical Method of application Rate of application Per hole Per acre Depth of appli- cation Plot No. Effectiveness expressed as per cent of initial count Of each plot Carbon disulfide Carbon tetrachloride Chloropicrin Carbon disulfide emulsion Calcium cyanide Equal parts of hy drated lime and sulfur Naphthalene. Lead arsenate Injected into the soil 18 inches apart each way Injected into the soil 18 inches apart each way Injected into the soil 18 inches apart each way Diluted 1 to 300; 5 gal. applied to the square yard Sown in furrows Broadcast and spaded under Broadcast and spaded under. Broadcast and spaded under fluid ounces 1.92 0.96 1.92 0.13 Equal parts of potas sium xanthate and sulfur Paradichloro- benzene Sown in furrows.. Sown in furrows. 290 gal 145 290 gal. 280 lbs. 300 lbs. 600 lbs. 900 lbs. 1,200 lbs. 1,200 lbs. 300 lbs. f 1,200 lbs. 2,400 lbs. 300 lbs. 600 lbs. 900 lbs. inches fe Applied to soil surface 2 1 26 15 :2b 16 \36 /> \20 18 37 4 21 6 23 14 38 28 33 19 124 17 35 1001 99f 100\ 100/ ioo\ 100/ 52) 66] 53) 54] 41) 11/ 33) —20*] 93\ 65] -20*"\ 20/ 53\ -35*( ■130*\ 5( 74] 1001 lOOf ioo\ 100/ * The minus percentages, indicating a greater relative increase of the centipede in these plots the check plots, are probably due to an underestimation of the initial population, as explained on than in page 12. 14 University of California — Experiment Station pounds per acre) the 93 per cent kill is undoubtedly high. The only ex- planation that can be offered is that the original population in this plot was low, and therefore, in the population studies before or after treat- ment a few centipedes observed one way or the other would tend to alter the results greatly. The same seems to be true of plot 6 (potassium xanthate + sulfur, 1,200 pounds) , for this material is unquestionably in- effective against the garden centipede. The minus percentage obtained in plot 14 (potassium xanthate + sulfur, 2,400 pounds per acre) is prob- ably due to an underestimation of the first population count, for several checks on the second count showed it to be correct. The other minus percentages are perhaps also the result of exceptionally low first counts, together with experimental error. From table 2 it is seen that the two chemicals producing near eradica- tion are paradichlorobenzene and carbon disulfide. Of these paradi- chlorobenzene was preferable when cost was taken into consideration. None of the other chemicals can be considered effective under the con- ditions of the experiment. Unless eradication or near eradication is ac- complished, chemical treatment against the garden centipede is not jus- tified. The one exception to this is carbon disulfide emulsion, and this will shortly be considered. Paradichlorobenzene. — Paradichlorobenzene, when used at the rate of 600 or 900 pounds per acre, gave, as far as could be ascertained, com- plete eradication of the centipede. Preliminary results had indicated that this chemical is not effective unless it is applied in the summer when the soil temperature is high, and unless thorough and even distribution is obtained. Packing the soil after it is applied is advisable because this prevents the evolving gas from escaping too easily from the soil. While this material is very effective against the garden centipede when applied under proper conditions, it has several disadvantages : (1) It should not be used where there is vegetation, for it is almost certain to be injurious. (2) The fumigant is rather slowly dissipated, and after its application several months should pass before treated ground is planted. (3) It should only be applied during the summer if satisfactory results are to be expected. Carbon Disulfide. — Carbon disulfide gave good results when used at the rate of 145 and 290 gallons per acre. However, when 145 gallons per acre are used, control can only be expected under the most favorable conditions. This is based upon rather extensive work other than that carried on in the work here described. As with paradichlorobenzene, best results can be expected in the summer when the soil is very warm. However, this material has a greater range of effectiveness than paradi- Bul. 548] Control of the Garden Centipede 15 chlorobenzene, and when used at the higher rate (290 gallons per acre) good results have been obtained during the spring and fall. This mate- rial kills with great rapidity, and in one plot examined 18 hours after treatment all animal life appeared to have been killed, including the garden centipede, true centipedes, sowbugs, earthworms, and millipeds. Several precautions should be used when treating soil with this mate- rial. After the material is applied the holes should be tamped tightly shut, and it is well to firm the whole soil surface by wetting or rolling to prevent the escape of the gas. At time of application the soil should not be too wet, as very wet soil apparently cuts down the effectiveness of the chemical by preventing the penetration of gas through it. In one greenhouse where a rather extensive area was treated with this chemical, very poor control was obtained, and for no apparent reason other than that the soil was very wet. Carbon disulfide is very inflammable and explosive when mixed with air. It should be kept away from all flame and from electric sparks, and smoking should be prohibited. This material is also injurious to growing plants, and, therefore, its use is mostly limited to areas free from vege- tation. However, it dissipates very rapidly, and planting can take place in a week after its application without injury. In fact, in some cases it has been reported that growth has been stimulated by it, probably owing to the partial sterilization of the soil. Carbon Disulfide Emulsion. — While carbon disulfide emulsion 8 gave unsatisfactory results when used in the experimental area, it has been tried in greenhouses with some success. This material has an advantage over paradichlorobenzene and carbon disulfide in that it can be used where vegetation is growing. When diluted 1 to 300, and applied at the rate of 5 gallons to the square yard, it has given fair to good control. At this rate of application, the chemical is not injurious to many plants, although some, such as asters, have been severely injured. This mate- rial is not effective unless the centipedes are near the soil surface, and the soil is wet, loose, and porous so that the liquid can penetrate it readily. Wetting the soil has a tendency to bring the centipedes close to the soil surface. In greenhouses where this material has been applied under opti- mum conditions, no living centipedes could be found the day after the treatments, although many dead ones were evident in every shovelful of soil examined. The diluted emulsion should be kept well agitated and applied soon after mixing to prevent the escape of the carbon disulfide s If carbon disulfide emulsion is mixed indoors, the room should be well ventilated. Further, the container in which it is mixed should be drained and washed. Otherwise, upon standing, there is some danger that enough carbon disulfide may escape from any remaining emulsion to form an explosive mixture. 16 University of California — Experiment Station gas, which is slowly given off. In experimental work, where rather large areas were to be treated, the dilute emulsion was made up in a 50-gallon barrel, and applied to the infested ground with a sprinkling can. It is very probable that any centipedes fairly deep in the soil will not be affected by the emulsion, and these, with others that are likely to migrate into the beds, will set up new infestations, making it necessary to treat the beds from time to time. This is particularly true where plants re- main in the soil for rather long periods. However, under favorable conditions a profitable crop of short-lived plants such as snapdragons, could probably be obtained with one or two applications of the emulsion. CONCLUSIONS From the work carried out it has been determined that paradichloro- benzene, carbon disulfide, and carbon disulfide emulsion are very effec- tive when used against the garden centipede. The rate at which these chemicals must be used in order to be effective will undoubtedly prohibit their use in the control of the centipede under field conditions, except in special cases. However, there are many greenhouses in which the centi- pede is proving to be a serious pest. In places it has made the production of certain plants uneconomical. In such places chemical treatment of the soil might readily relieve this situation. The only other alternative is to install raised benches where the plants to be grown can be propa- gated in this type of bed. SUMMARY In order to study the effect of chemical treatment in combatting the garden centitpede, Scutigerella immaculata (Newport), experimental plots were isolated by keeping them continuously surrounded by water. Chemicals experimented with were paradichlorobenzene, calcium cyanide, naphthalene, lead arsenate, potassium xanthate mixed with equal parts of sulfur, hydrate of lime mixed with equal parts of sulfur, carbon tetrachloride, carbon disulfide, carbon disulfide emulsion, and chloropicrin. The experimental plots were treated during the summer when the soil temperature was high, and this is believed to have allowed the chemicals to exert their fullest lethal power. During the course of the experiment soil temperatures were recorded for 3 and 10-inch soil levels, a Bristol's recording thermometer being used for this purpose. In order to determine the effectiveness of the various chemicals, popu- lation studies were made before and after treating the plots, and Bul. 548] Control of the Garden Centipede 17 effectiveness calculated by a formula to correct for a change in the method of counting population and for normal increase in numbers. Examinations after treating the plots showed that carbon disulfide, carbon disulfide emulsion, carbon tetrachloride, chloropicrin, and cal- cium cyanide killed rapidly, with the first three, at least, exerting their fullest killing power during the first 24 hours. Paradichlorobenzene and naphthalene appeared to kill rather slowly, with the lethal action of paradichlorobenzene lasting until apparently all the centipedes are killed. Of the chemicals experimented with, the only ones that showed prom- ise in the control of the garden centipede were paradichlorobenzene, carbon disulfide, and carbon disulfide emulsion, when the rate of appli- cation was heavy enough. While carbon disulfide emulsion did not give good results when used in the experimental plots, it has given satisfactory control under certain greenhouse conditions. It is effective only when the centipedes are con- centrated near the surface of the soil, and the soil is loose, porous, and moist so that the emulsion penetrates it readily. Paradichlorobenzene proved effective when used at the rate of 600 and 900 pounds per acre. This chemical should be applied only in the summer when the soil temperature is high, if effective control is to be obtained. Carbon disulfide proved effective when used at the rate of 145 and 290 gallons per acre. The effectiveness of the 145-gallon treatment is believed to be due to the fact that it was applied under most favorable conditions. Paradichlorobenzene and carbon disulfide are both injurious to vege- tation and, therefore, their use is mostly limited to areas free of growing plants. Carbon disulfide emulsion has an advantage over the above-mentioned chemicals in that it can be applied to many different kinds of plants without causing injury to most of them. Because of cost, chemical control of the centipede under field condi- tions will probably prove uneconomical in most cases, although on small areas and in greenhouses such treatment may be profitable. ACKNOWLEDGMENTS I am greatly indebted to Professors Herms, Essig, and Hoskins, and to F. H. Wymore, for many helpful suggestions in conducting this in- vestigation. I also wish to express my thanks to J. Holmes and Sons of Clarksburg who furnished land and labor for the carrying out of the experimental work. 18 University of California — Experiment Station LITERATURE CITED i Compton, Charles C. 1930. Greenhouse pests. Illinois Nat. Hist. Survey Ent. Ser. Cir. 12:1-112. 2 Essig, E. O. 1926. Paradichlorobenzene as a soil fumigant. California Agr. Exp. Sta. Bui. 411:1-20. 3 Feytaud, J. 1926. Contribution a 1 'etude de symphyles. Rev. Zool. Agr. et Appl. [Bor- deaux] 25 (1 and 2):1-11; 25-32. Reviewed in: Rev. Appl. Ent. Ser. A 14, 1926:328. 4 FlLINGER, G. A. 1928. Observations on the habits and control of the garden centipede, Scuti- gerella immaculata, a pest in greenhouses. Jour. Econ. Ent. 21(2): 357-360. 5 FlLINGER, G. A. 1931. The garden symphylid, Scutigerella immaculata Newport. Ohio Agr. Exp. Sta. Bui. 486:1-33. e Herrick, G. W. 1927. Scutigerella immaculata, a pest in greenhouses. Jour. Econ. Ent. 20(5) : 738-740. 7 Illinois Agricultural Experiment Station. 1930. A year's progress in solving farm problems in Illinois. Illinois Agr. Exp. Sta. Ann. Rept. 1929-30:149. s McDaniel, E. I. 1931. Insects and allied pests of plants grown under glass. Michigan Agr. Exp. Sta. Special Bui. 214:1-117. 9 Oregon Agricultural Experiment Station. 1926. Report of the Department of Entomology of Oregon Station. Oregon Agr. Exp. Sta. Bien. Rept. 1924-26:96-107. io Oregon Agricultural Experiment Station. 1928. Report of work with economic insects and their control at the Oregon Station. Oregon Agr. Exp. Sta. Bien. Rept. 1926-28:101-109. ii Riley, H. K. 1928. A method of volatilizing carbon disulfide for soil fumigation. Jour. Econ. Ent. 21(6) :932-933. 12 Riley, H. K. 1929. The greenhouse centipede. Purdue University, Indiana Agr. Exp. Sta. Bui. 331:1-14. is Sheaffer, Frank E. 1930. Some insect pest and plant diseases of Indiana. Indiana Dept. Conserv. Div. Entom. Pub. 103:54-55. Bul. 548] Control of the Garden Centipede 19 14 Thomas, C. A. 1928. Control of millipedes. In: Pennsylvania Agr. Exp. Sta. 41st annual report. Pennsylvania Agr. Exp. Sta. Bul. 230:40. is Wallace, F. N., et al. 1926. Eeport of the Division of Entomology. Indiana Dept. Conserv. Ann. Kept. 1925:53. is Walton, C. L. 1930. Notes on serious tomato pest {Scutigerella immaculate). Univ. Bristol Agr. and Hort. Res. Sta. Ann. Eept. pp. 76-78. Long Ashton, Bristol, England. 17 Wymore, F. H. 1924. The garden centipede, Scutigerella immaculata (Newp.), a pest of eco- nomic importance in the West. Jour. Econ. Ent. 17(5):520-526. is Wymore, F. H. 1931. The garden centipede. California Agr. Exp. Sta. Bul. 518:1-22. 12m-12,'32