i 953 
 16 
 
 >py 1 
 
 Spreading and Adherence 
 of Arsenical Sprays 
 
 A THESIS 
 
 Presented to the Faculty of the Graduate School 
 OF Cornell University for the degree of 
 
 DOCTOR OF PHILOSOPHY 
 
 BY 
 WILLIAM MOORE 
 
 Reprinted from University of Minnesota Agricultural Experiment Station 
 Technical Bulletin 2, June, 1921 
 
spreading and Adherence 
 of Arsenical Sprays 
 
 A THESIS 
 
 Presented to the Faculty of the Graduate School 
 OF Cornell University for the degree of 
 
 DOCTOR OF PHILOSOPHY 
 
 BY 
 
 WILLIAM MOORE 
 
 Reprinted from University of Minnesota Agricultural Experiment Station 
 Technical Bulletin 2, June, 1921 
 
^^ 
 
 
 TABLE OF CONTENTS 
 
 Pasre 
 
 Summary ' 
 
 Introduction ^ 
 
 Definition of terms 5 
 
 Historical consideration of spreading 5 
 
 The theory of spreading 15 
 
 Sprays that will spread on cabbage leaves 18 
 
 Sprays that will spread on other leaves 20 
 
 Conclusion regarding the phenomenon of spreading 22 
 
 Historical consideration of adherence 22 
 
 Use of organic materials with fungicides 22 
 
 Use of organic materials with insecticides 24 
 
 Use of inorganic substances 26 
 
 A new theory of adherence 29 
 
 Electric charge exhibited by the common arsenical preparations 30 
 
 Electric charge exhibited by the leaf surface 31 
 
 Preparation of an electrically positive arsenical substance 32 
 
 Field tests of adherence 34 
 
 Preparation of positively charged ferric arsenate 36 
 
 Comparative toxicity of different arsenical preparations 39 
 
 Methods of determining comparative toxicity 40 
 
 Comparative tests of toxicity 41 
 
 A method of expressing toxicity 43 
 
 Influence of ferric hydroxide on toxicity 43 
 
 Conclusion regarding the phenomenon of adherence 44 
 
 Literature cited 46 
 
i. 
 
 SPREADING AND ADHERENCE OF ARSENICAL 
 
 SPRAYS 
 
 By William Moore 
 
 - SUMMARY 
 
 The addition of material similar in chemical constitution to the leaf 
 surface causes the spray mixture to form a film of liquid over the leaf. 
 
 The positive adsorption of the added material at the leaf-spray inter- 
 face, resulting in a lowering of the interfacial tension, appears to offer 
 the best explanation of the results. 
 
 Different types of leaves naturally require different materials. Thus 
 organic compounds such as beechwood creosote, carvacrol, or amyl alco- 
 hol, soluble in fats and waxes and but slightly soluble in water, produce 
 good spreading over waxy leaves, such as cabbage. 
 
 Various proteins and plant infusions give good spreading on leaves 
 with surfaces of cellulose, even when they are strongly cutinized, as in 
 the case of plum and citrus leaves. 
 
 Suspensions containing small-sized particles adhere better than those 
 with larger particles. 
 
 An even distribution of the spray over the leaf tends to increase the 
 adherence. 
 
 The leaf surface, when wet, exhibits a negative electric charge. 
 
 The common compounds of arsenic, such as lead arsenate, paris 
 green, calcium arsenate, and others, have particles carrying negative 
 electric charges. 
 
 Arsenic compounds of aluminum, chromium, and iron may be pre- 
 pared so that the particles carry a positive charge. 
 
 Ferric arsenate appears the most promising and is more toxic than 
 lead arsenate. 
 
 Field tests show that electrically positive arsenical preparations ad- 
 here more strongly to the leaf surface than do those which are negatively 
 charged. 
 
 Ferric oxide, or hydroxide, by adsorbing compounds of arsenic, 
 lower their toxicity to insects. 
 
 The ratio of the amount of the arsenic compound in the body to that 
 in the excreta is a better basis of comparing toxicity of different arsen- 
 ical preparations than tests based on the food consumed or the time re- 
 quired to produce death. 
 
 1 The writer wishes to acknowledge the assistance of Dr. R. A. Gortner, of the Uni- 
 versity of Minnesota, on certain phases of the problem, particularly the studies of cataphoresis 
 and endosmosis. 
 
TECHNICAL BULLETIN 2 
 
 INTRODUCTION 
 
 The successful use of arsenical sprays for the control of various in- 
 jurious insects depends upon a number of factors. The simple recom- 
 mendation of paris green or lead arsenate is not sufficient, since the re- 
 sults will vary greatly according to the character of the plant or insect. 
 If the plant is particularly sensitive to arsenic, care must be taken to 
 use an arsenical material with a very low percentage of soluble arsenic. 
 In spraying plants subjected to moist conditions resulting from frequent 
 light rains or fogs, the stability of the arsenical preparation becomes of 
 prime importance. Many compounds of arsenic are decomposed by the 
 action of water and carbon dioxide. 
 
 Leaves require a spray mixture of such a character that the film pro- 
 duced at the time of spraying will persist and will not tend to collect 
 in drops and roll ofT. This is particularly true of waxy leaves, such 
 as those of cabbage, or strongly cutinized leaves, such as those of the 
 citrus fruits. Such spray mixtures are not only more economical but 
 are also of great importance in the control of insects, which, like the 
 plum curculio, feed on isolated portions of the plant. 
 
 Not only the nature of the plant but also that of the insect must be 
 considered. Some insects are easily destroyed by arsenical poisons, 
 while others require large qtiantities to produce the desired results. 
 When the plant to be treated is not sensitive to soluble arsenic, a resist- 
 ant insect may be destroyed by the use of a compound containing a rather 
 high percentage of free arsenious oxide, or one which readily decom- 
 poses with the liberation of soluble arsenic. In most cases, however, the 
 use of a very stable and insoluble arsenical substance is required. Only 
 a small percentage of such a poison is absorbed by the alimentary tract 
 of some insects, the major portion being excreted. The excreta in such 
 cases may be white in color owing to the presence of lead arsenate, and 
 chemical analysis will frequently show that more than fifty per cent of 
 the poison consumed was not absorbed. By greatly increasing the quan- 
 tity of the arsenical material applied to the leaves, the death of such in- 
 sects may be produced, but such an increase is not always justified from 
 an econoriiic standpoint. Smaller quantities of the poison may be suc- 
 cessfully used, providing it remains on the leaves long enough for the 
 insect to consume a killing dose. The adherence of the dried material 
 of the spray, that is, its ability to withstand the washing effect of rain 
 and dew, thus assumes much importance. Plants subject to the attack 
 of an insect over a long period of time must be treated with a very ad- 
 herent spray, or the treatment must be repeated several times. 
 
 The present investigation is a study of the principles governing the 
 uniform distribution of the spray material over the plant, and the adher- 
 ence of the dried particles when subjected to rain, dew. wind, and other 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 5 
 
 influences. The object of the study is general, and no effort has been 
 made to apply the results to the control of any particular insect. 
 
 DEFINITION OF TERMS 
 
 Many attempts have been made, in the use of insecticides and fungi- 
 cides, to insure even distribution of the spray over the foliage of the 
 treated plant. In many of the papers dealing with this matter it is not 
 clear whether the investigator is considering the spreading of the spray 
 over the foliage or the adherence of the dry materials. Statements that 
 the spray did not "stick" or adhere well to the plant are frequent. In 
 some cases, from the context, it is apparent that reference is made to 
 spreading, the drops of the spray having failed to adhere to or spread 
 over the leaves. The term "wetting" is frequently used as a synonym of 
 spreading. Throughout this paper the word "spreading" will be used 
 to denote the formation or the maintenance after being formed, of a 
 continuous film over the surface of the leaf, and "adherence" will apply 
 to the resistance to the action of rain, dew, and wind exhibited by the 
 spray material after it dries. The term "wetting" may be considered as 
 the slight chemical or physical affinity between the liquid and the solid, 
 which is one of the factors in the formation of a continuous film. 
 
 HISTORICAL CONSIDERATION OF SPREADING 
 
 One of the earliest records of the use of a material which tends to 
 produce the spreading of a spray over the leaves is contained in the 
 Rapport au Ministre de L' Agriculture (1885). This report deals with 
 the treatment of the mildew of the vine, and Davis, an investigator, is 
 reported to have used 6 kilos of glue to 800 liters of copper sulphate 
 solution, thinking that the glue increased the efficacy of the remedy. 
 Whether the increased value of the spray was caused by a more even 
 spreading or by a better adherence is not mentioned. The following year, 
 Millardet and Davis (1885) said that the addition of glue to bordeaux 
 did not appear to be of notable advantage. 
 
 Soap or soapsuds was early used in contact sprays with tobacco and 
 other materials in order to insure the spreading of the insecticide over 
 the body of the insect. Altho the earliest applications of paris green 
 were in the form of dust, the use of suspensions of paris green or Ion- 
 don purple in water soon developed. Gillette (1890) used soapy solu- 
 tions with paris green and london purple, but found that these mixtures 
 were more injurious than a suspension containing no soap. The addition 
 of flour at the rate of half an ounce to a gallon was- tried, but this 
 preparation also proved more injurious. The increased injury, it was 
 thought, was due to the greater adhesiveness of the mixture. Resin in 
 the form of a soap, sodium resinate. was successfully used without in- 
 creasing the amount of injury to the foliage. Washburn (1891) used 
 
6 TECHNICAL BULLETIN 2 
 
 whale oil soap, 6 pounds to 50 gallons of a paris green suspension, 
 thereby producing an even spread of the mixture over the fruit and 
 leaves, and apparently rendering the poison more tenacious. Galloway 
 (1892) used soap with different fungicides to increase spreading. 
 Swingle (1894) recommended the use of soap with bordeaux mixture. 
 No definite quantity was specified, it was said merely that sufficient soap 
 should be added to give a good permanent foam. This result was usu- 
 ally obtained by using a quantity of soap equal to one half the com- 
 bined weight of the copper sulphate and lime in the bordeaux mixture. 
 Fairchild(i894)used soap with eau celeste, obtaining excellent spreading 
 even on leaves with a heavy cuticle, such as those of the pear. Galloway 
 (1894) used a resin soap with bordeaux mixture, finding it just as 
 effective in producing a continuous film on the leaf as Ivory or whale oil 
 soap. Lowe (1896) encountered considerable difficulty in spreading a 
 lead arsenate spray evenly over the surface of willow leaves. Glue used 
 at the rate of 2 quarts to 45 gallons of spray gave satisfactory results. 
 
 The indefiniteness encountered in the use of soap as a spreader for 
 various spray mixtures was shown by the analysis of various soaps by 
 Van Slyke and Urner (1904). They showed that in the different makes 
 of whale oil soap the water content varied from 11. 15 to 54.85 per cent 
 and the amount of actual soap from 14.90 to 59.27 per cent. The nature 
 of the alkali present varied greatly, as did also the quantity of fatty 
 acids and resins. Other common soaps showed a similar variation. The 
 addition, therefore, of a definite amount of soap to a spray may or may 
 not cause an even distribution of the material over the leaves. 
 
 Alausier (1908) studied the spreading of a number of liquids over 
 different solids. He realized that spreading is dependent on the nature 
 of the solid and of the liquid, but considers this difference as due to the 
 surface tension of the liquid. A liquid which will spread over a par- 
 ticular solid must possess a certain definite surface tension value. If 
 it has this value it will spread regardless of the nature of the liquid. 
 A spray containing 30 grams of soap to 10 liters of water with either 50 
 grams of oil of tar or 10 grams of formaldehyde is considered as meet- 
 ing the surface tension requirements of the leaf. 
 
 Vermorel and Dantony (1910) state that a body is wetted by a 
 liquid, i.e., the liquid spreads over it when the cohesion of the molecules 
 of the liquid, the one for the other, is less than double their adhesion for 
 the solid. They distinguished between the surface tension of a fresh 
 surface of a soap solution, dynamic surface tension, and the surface 
 tension of an old surface, static surface tension, considering that there 
 was a direct relationship between the static surface tension of a soap 
 solution and its ability to spread over an insect. Measurements of the 
 surface tensions, by means of a stalagmometer, of several organic 
 compounds were made, and it was found that none possessed as low a 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 7 
 
 surface tension as a soap solution. Since no mention is made of the 
 time allowed for the formation and breaking away of the drop from the 
 tip of the stalagmometer, it is doubtful whether sufficient time was al- 
 lowed for the concentration of the soap in the surface layer, thus insur- 
 ing a true measurement of the static surface tension. In a second paper, 
 Vermorel and Dantony (1911) take up the addition of soap to fungi- 
 cides to increase their spreading, showing^ that the results obtained de- 
 pend largely upon the method used in the preparation of the mixture. 
 A method is mentioned for the preparation of a colloid copper soap 
 which exhibited good spreading properties. 
 
 Parker (1911) found that the addition of 2 or 3 bars of common 
 laundry soap to each 50 gallons of a lead arsenate spray retarded the 
 settling out of the suspended particles. The use of soap with arsenical 
 sprays to retard settling and to increase spreading, when tried under or- 
 chard conditions by certain growers, was frequently accompanied by an 
 increased injury to the foliage of the sprayed plants. These results 
 caused Tartar and Bundy (1913) to investigate the action of soap on 
 lead arsenate. They showed by chemical analysis that the addition of 
 different soaps had little influence on the neutral lead arsenate 
 Pb3(As04)2, but that the addition of soap to the acid lead arsenate 
 PbHAs04, rendered soluble a considerable quantity of the arsenate. 
 Since the acid lead arsenate is more generally used than the neutral, the 
 addition of soap to cause spreading can not be safely recommended. 
 Edwardes-Ker (1913) also studied the addition of soap to lead arsenate 
 sprays, obtaining results opposed to those of Tartar and Bundy. He 
 found that the addition of i per cent of soap, either home-made or com- 
 mercial, did not influence the solubility of lead arsenate suspensions. 
 Both home-made lead arsenate, formed by the reaction of lead acetate 
 and sodium arsenate, and commercial lead arsenate were used in the ex- 
 periment. No distinction is drawn between the difi^erent kinds of lead 
 arsenate. The mixtures of lead arsenate and soap were filtered half an 
 hour" after being made up, and the arsenic in the filtrate was determined 
 gravimetrically as magnesium pyroarsenate. Tartar and Bundy used 
 less soap, 4.8 grams to 1,000 cubic centimeters of water, but the mixture 
 was allowed to stand six hours before filtering. They used the iodine 
 titration method of Gooch and Browning in the determination of the 
 quantity of soluble arsenic. As the lead arsenate and soap would gen- 
 erally be mixed for periods longer than half an hour, and even after 
 spraying would have an opportunity for chemical reactions, it appears 
 that the results of Tartar and Bundy more nearly represent the effect of 
 adding soap to lead arsenate sprays. 
 
 Gastine (1912) recommended the use of saponin to reduce the sur- 
 face tension of spray materials and hence to insure spreading. One of 
 its advantages is said to be that, unlike soap, it is not precipitated by acids 
 
8 TECHNICAL BULLETIN 2 
 
 or metallic solutions, and will not cause injury to the plants. Saponin 
 can therefore be added to either fungicides or insecticides without dan- 
 ger of any decomposition resulting. It may be used instead of soap in 
 the formation of emulsions. Used with beechwood creosote or coal-tar 
 creosote the surface tension of the mixture was lower than that of the 
 pure saponin solution. 
 
 Weinmann (191 2) prepared a cupric spray with calcium poly sul- 
 phide and soap, finding that this combination gave a very high drop num- 
 ber and spread well. The drop number of the cupric spray with soap, 
 but without the polysulphide, was much lower. The substitution of 
 saponin in place of soap also gave inferior results. 
 
 Vermorel and Dantony (1912), continuing their investigations, con- 
 cluded that the surface tension of a spray solution is not an index of 
 its spreading power. Solutions with different surface tensions may 
 have the same ability to spread, or the spreading power may vary ac- 
 cording to the nature or physical state of the material to be treated. 
 Two distinctions are made : ( i ) Where the liquid is able to touch the 
 surface, when it may spread as water on a potato leaf or may collect in 
 a drop as water on a vine (grape) leaf; (2) where the liquid rests on 
 the surface without touching it, as in the case of a drop of water on a 
 cabbage leaf, a film of air being interposed between the leaf and the 
 water. A solution of sodium oleate giving a stalagmometer reading of 
 142 drops may spread over a cabbage leaf and fail to spread on the leaf 
 of a vine. They believed that the surface tension of the spray in- 
 fluenced its spreading less than did its surface viscosity. Whereas the 
 sodium oleate solution giving 142 drops from the stalagmometer failed 
 to spread over the leaves of the vine, a saponin solution giving only loi 
 drops, and therefore apparently possessing a higher surface tension, 
 spread readily. This difiference is thought to be explained by the 
 higher surface viscosity of the saponin solution, so that when spread in 
 a thin film over the leaf, surface concentration rapidly took place, thus 
 increasing the surface viscosity to such an extent that the solution was 
 unable to collect in a drop. A gelatin solution of i part in 10,000 parts 
 of water, possessing a surface tension nearly equal to water, caused 
 spreading when applied to the leaves of a vine. Gelatin in solution was 
 more effective than either saponin or soap and had the further advantage 
 of being uninfluenced by either acid, neutral, or basic materials. 
 
 In this as in the preceding paper (Vemiorel and Dantony, 1910), no 
 mention is made of the speed of formation or the rate of detachment of 
 the drops from the tip of the stalagmometer. Concentration of the ma- 
 terials in the surface layer reduces the surface tension, and since dif- 
 ferent materials may vary as to the time required to concentrate in the 
 surface layer, these measurements of surface tension are not comparable. 
 Further, if surface concentration takes place so rapidly as to cause the 
 
SPREADING AND ADHERENCE OF ARSENIC AE SPRAYS 9 
 
 film to become so viscous that it can no longer collect in a drop, a de- 
 cided lowering of the surface tension should be apparent in the stalag- 
 mometer readings. No consideration is taken of the possibility of a 
 concentration of the materials at the interface of leaf and spray which 
 would result in a lowering of the surface tension at the interface and 
 tend to produce spreading. 
 
 Chappaz, G.. (1913) recommends the use of saponin in preparing 
 sprays which will spread well and considers it superior td soap. Chapaz 
 (1913) reviews the recent work with spreaders, discussing soap, gelatin, 
 glue, and saponin. He recognizes a general but not a mathematical rela- 
 tionship between spreading and the drop number determined by means 
 of a Duclaux pipette ( stalagmometer ) . Vermorel and Dantony (1913) 
 recorded several methods of preparing fungicides with casein. A solu- 
 tion of casein is prepared by dissolving it in a solution of 100 grams of 
 sodium carbonate to a liter of water. Parker (1913) used from 4 to 10 
 pounds flour paste with lime sulphur and nicotine sulphate to each 100 
 gallons of liquid. The spray was used primarily for red spider and was 
 tried on several dififerent plants. Altho it was an effective spreader on 
 many plants, it failed when used on carnations, sweet peas, and green- 
 house roses. The possibility of its use as a "sticker" for lead arsenate 
 suspensions was considered. Glue and miscible oil mixed with lime sul- 
 phur was used for the control of red spider by Jones (1913). Excellent 
 spreading was reported on all parts of plum and peach trees and it is be- 
 lieved that similar results would be obtained on citrus trees. Neiils 
 (1913) confirmed the results of Parker (1913) on the spreading prop- 
 erties of lime sulphur and flour paste and the value of this spray for the 
 destruction of red spider. 
 
 Lafforgue (1913) points out that spreading and adherence are two 
 entirely different phenomena. The addition of saponin to the spray 
 mixture gives good spreading but not good adherence, while gelatine 
 spreads well and also improves the adherence. 
 
 Lefroy (1913), working in India, had determinations made of the 
 surface tension of the solutions used for controlling psylla attacking in- 
 digo. Soap solutions showed a very much lower surface tension than 
 water, bu^ soap emulsions of oil or creosote exhibited a higher surface 
 tension than the soap solution alone. These results are opposed to the 
 work of Gastine (1912), who found that a saponin emulsion of beech- 
 wood creosote or coal-tar creosote possessed a lower surface tension than 
 the saponin solution. However, since no mention is made by Lefroy of 
 the method employed in determining the surface tension, and in view of 
 the fact that in one case soap and in the other saponin was used to emul- 
 sify the creosote, no conclusion can be reached as to the accuracy of the 
 results. 
 
10 TECHNICAL BULLETIN 2 
 
 Astruc (191 3) divides spreaders into four types: (i) Soaps; (2) 
 saponins; (3) albuminous substances, such as albumin, gelatin, milk, 
 dried blood, and casein; (4) organic salts of more or less limited pro- 
 duction. No mention is made as to the nature of these organic salts. 
 Issleib (1914) extracted 2 kilograms of "carrangeen," or pearl moss 
 with 100 kilograms of water. The extract, a thick slimy liquid, when 
 sprayed over the leaves formed a thick continuous coating which on 
 drying cracked and scaled off, carrying with it the larvae or eggs of in- 
 sects infesting the plant. These results are comparable to those re- 
 corded by Parker (1913), using flour paste, and no doubt this prepara- 
 tion would give similar spreading. 
 
 Del Guercio (1914), in Italy, recommended the use of i to 2 per 
 cent of flour paste, fish glue, or other colloidal substance, with potas- 
 sium or calcium polysulphide to give an even distribution of the spray 
 over citrus trees to control Chrysoniphalus dictyospcrnii var. pinnuli- 
 fcra. Rye flour is preferred, but spoiled wheat flour or the flour of 
 other grains may be used. An even distribution was obtained regard- 
 less of the degree of maturity or the position of the treated portions. 
 Arnal (1914) mentions that "soluble caseins," for use with bordeaux 
 mixture, are being manufactured in France and put on the market. It 
 may be that these preparations are similar to the casein glue being sold 
 in the United States. 
 
 Vermorel and Dantony (191 5) prepared rapidly a calcium caseinate 
 by stirring a liter of milk of lime into a suspension of 100 grams of 
 casein powder in a liter of water. The milk of lime may contain from 
 50 to 200 grams of lime without influencing the results. One liter of 
 this solution is added to a hectoliter of alkaline bordeaux mixture to 
 give spreading and adherence. 
 
 The most important paper, following the papers of Vermorel and 
 Dantony, dealing with the spreading of sprays, is that of Lefroy 
 (191 5). The fact that spreading depends not only upon the value of 
 the surface tension of the spray, but also upon the surface tension of 
 the leaf^ and the surface tension at the interface of the spray and the 
 leaf, is presented by Lefroy for the first time in entomological literature. 
 Unfortunately Lefroy becomes confused in his terms, and draws the 
 incorrect conclusion that if the surface tension at the interface between 
 the spray and the leaf is greater than the sum of the surface tension 
 of the spray and that of the leaf, spreading will occur. Realizing that 
 the surface tension of the leaf and the surface tension at the interface 
 of leaf and spray can not be readily determined, he recommends the 
 determination of the surface tension of the spray, which should be as low 
 as possible. Using the term "wetting" to denote the specific attraction 
 
 ^ Solids are usually considered as possessing a surface tension altho it has not been 
 demonstrated. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS n 
 
 or slight chemical affinity existing between the spray and the leaf, the 
 statement is made that "spreading of a liquid over a surface it never 
 wets" may occur. This conclusion does not appear to be justifiable, 
 since if no wetting occurred the surface tension at the interface would 
 be so high that spreading would be impossible. The theory of spread- 
 ing of both sprays and dips is presented in a very interesting paper by 
 Cooper and Nuttall (1915). First pointing out the value of having 
 sprays and dips which will spread over the treated plant or animal, 
 they give the principles of spreading established by Quincke ( 1877 K 
 Quincke studied the capillary rise of liquids in tubes of different ma- 
 terials, the spreading of one liquid over another liquid or a solid, and 
 the angles formed by the liquid at its point of contact. From this 
 study he formulated the conditions necessary for spreading to occur. 
 These conditions in terms of sprays may be expressed by the statement 
 that if the surface tension of the solid (leaf) is greater than the sum 
 of the surface tension of the spray and the surface tension at the inter- 
 face of spray and solid, spreading will occur. Cooper and Nuttall 
 believe that in general this statement holds, but believe that there 
 may be certain exceptions. Two factors are thought to be able to upset 
 entirely this relationship. The first of these is the solvent action of the 
 liquid on the solid, particularly in those cases where the solid is coated 
 with wax or grease. This solvent action of the spray is believed to be 
 independent of the surface tension at the interface of solid and liquid, 
 but as will be shown later in this paper the solvent action of one liquid 
 on another liqviid or a solid, results in the lowering of the surface ten- 
 sion at the interface. Sprays, therefore, which spread because of their 
 solvent action, are not exceptions but actually fulfil the surface tension 
 conditions necessary for spreading to occur. 
 
 The second factor considered as vitiating the surface tension equa- 
 tion is the phenomenon of surface concentration. The experiments of 
 Vermorel and Dantony are cited as showing that a spray with a high 
 surface tension spreads well owing to its high surface viscosity resulting 
 from a rapid surface concentration of the solute. 
 
 Since these experiments have already been discussed and will be 
 taken up later they will not be considered further here, tho it may be 
 said that in the opinion of the writer, they do not invalidate the surface 
 tension relations already expressed. 
 
 Having presented the problem. Cooper and Nuttall take up the possi- 
 bility of measuring the spreading power of dips and sprays. The work 
 of Vermorel and Dantony is thought to show that the measurement 
 of surface tension alone is insufficient to determine the spreading quali- 
 ties of the liquid. The ability of the spray to produce a foam, since it 
 depends upon surface tension and does not express the surface tension 
 at the interface of the liquid and solid, can not be used as an index to 
 
12 TECHNICAL BULLETIN 2 
 
 spreading. The weight of the quantity of hquid adhering to a soHd 
 dipped into it is largely influenced by viscosity, and is therefore not 
 considered an accurate method of determining spreading". Spreading 
 may be conveniently expressed by the formula b s> b 1+ b Is where " ^ " 
 represents surface tension; "s", solid; "1", liquid; and "Is", the liquid- 
 solid interface. Then bs — ( bl+bls)= the spreading power of the 
 liquid expressed in dynes. Cooper and Nuttall, being chiefly inter- 
 ested in dips where the solid is represented by the skin of the animal, 
 usually oily in character, adopted castor oil or liquid vaseline as a 
 liquid substitute for the solid. This substitution simplified the prob- 
 lem, as the surface tension of the oil and of the liquid (dip) could easily 
 be measured and expressed in dynes. Using a modification of the drop 
 pipette method for determining the emulsifying action of soap solutions, 
 they measured the surface tension at the interface of the oil and the 
 dip. Substituting the values of the different surface tensions in the 
 above formula, a definite value for spreading, expressed in dynes, was 
 obtained. It is pointed out, however, that this method is applicable 
 only to soap solutions and will not give results with gelatine, saponin, 
 or similar solutions. 
 
 This method, based on the use of an oil as a substitute for the 
 solid, may be used only when the solid is oily or greasy, and can not 
 be used to determine the spreading power of sprays on leaves which 
 possess an entirely dififerent surface. 
 
 From the results of their experiments they conclude that for com- 
 parative values the interfacial tension alone may be considered. Since 
 this is inversely proportional to the number of drops produced by a 
 given volume of the oil, and since the spreading varies inversely as the 
 interfacial tension, it is directly proportional to the drop number. This 
 conclusion, however, is based on the assumption that the liquid has a 
 low surface tension, which, altho correct in the case of soap solutions, 
 is not always true of spray materials which will spread. Of the two 
 surface tensions, they consider the surface tension of the solid-liquid 
 interface more important. 
 
 A new material and one well adapted for use in certain countries 
 was suggested by High (1915). Fifteen pounds of cactus leaves sliced 
 and soaked in 50 gallons of water over night is recommended as a 
 spreader. Cactus extract gave excellent results when used with zinc 
 arsenite, and paris green and lime, but with lead arsenate the results 
 were unfavorable, owing to the formation of a precipitate. Cactus 
 grown near water contains a lower percentage of the desired material 
 and hence larger quantities were required to give the desired results. 
 
 Smith, L. B. (1916), using the method of finding the surface ten- 
 sion at the liquid-solid interface suggested by Cooper and Nuttall, deter- 
 mined the spreading power of various combinations of fish oil soap 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 13 
 
 and nicotine sulphate. No efifort was made to determine the surface 
 tension of either the oil or the spray. The sprays were tested with 
 AcyrtJwsiphmn pisi Kalt. on garden peas, Myciis persicac Sulz. on spin- 
 ach, and Tcfranychus sp. on strawberries. The maximum killing was 
 reached with certain concentrations of soap or soap and nicotine sul- 
 phate. Further additions of soap failed to increase the efficacy of the 
 spray. Using a fixed concentration of soap and increasing the concen- 
 tration of nicotine sulphate reduced the spreacling and also the killing- 
 power of the spray. This result was undoubtedly due to the reaction 
 between nico:ine sulphate and soap. 
 
 Lees (1917) tested the spreading of paraffin (kerosene) emulsion 
 on cloth surfaces previously painted with boiled linseed oil and dried. 
 A drop of the emulsion was rolled about on this surface and arbitrary 
 values from one to three were used to express the degree of spreading 
 of the kerosene and similar values for the aqueous phase. Based upon 
 these preliminary tests, further experiments were made on the spread- 
 ing of water solutions of gelatin, casein, soap, and paraffin emulsions 
 on the leaves of gooseberry and sea kale and on goosebrry mildew. All 
 the preparations gave complete spreading on gooseberry leaves buc only 
 certain of the emulsions gave complete spreading on kale leaves and 
 gooseberry mildew. No attempt is made to explain the results. 
 
 Altho not directly bearing on the spreading of sprays on foliage, 
 the experiments of Gray (1918) in preparing "wettable" sulphurs are 
 of interest. An ounce and a half of powdered glue dissolved in 3 gal- 
 lons of water thoroly wets, or spreads over, powdered sulphur, thus 
 making possible the formation of a suspension. 
 
 Lovett (1918) studied spreaders for arsenate sprays. Spreading is 
 thought to be caused largely by surface tension and specific gravity. 
 It is stated that the ability of the liquid to hold the arsenate in suspen- 
 sion is a very fair indication of its ability to spread. This is not al- 
 ways true, for altho some spreaders which increase the viscosity of the 
 medium also reduce the rate at which the suspended particles settle, this 
 characteristic in itself can not be considered an index of spreading. 
 Soap, glycerine, glue, casein, and an infusion of sage were tes'ed as to 
 spreading ability and also as to the degree of injury they produced on 
 bean foliage. An infusion of sage made by steeping Artcmcsia trUcn- 
 fata, Nutt, for twelve hours gave very encouraging results. The oils 
 and other ingredients present in the infusion were not determined but 
 were considered as probably similar to those found in Artemcsia frigida, 
 Willd. by Robak (1906). This species was shown to contain an es- 
 sential oil composed primarily of borneol camphor and cineol. From 
 4 to 8 ounces of casein converted into calcium caseinate and added to 
 each 100 gallons of spray gave good spreading. 
 
M TECHNICAL BULLETIN 2 
 
 Stearns (1920) experimented with the "Irish" moss which is simi- 
 lar to that used by Issleib (1914), as a spreader of arsenical substances 
 over peach foliage. Issleib's proportion of 4 pounds to 20 gallons was 
 found to be too thick for spraying purposes, but when used at the rate of 
 4 pounds to 50 gallons gave good results. Calcium caseinate at the rate 
 of I pound to 50 gallons also gave good results on peach foliage. 
 
 The most recent paper dealing with spreaders is that of Lovett 
 (1920). In discussing the suspension test as an index of spreading, he 
 says that altho not a definite criterion of spreading it "does indicate a 
 physical quality in the solution much to be desired in a spreader." His 
 belief that increased viscosity is accompanied by increased spreading is 
 shown in the experiments where phosphates and sulphates are added 
 to aqueous solutions of glue and of calcium caseinate to increase the 
 viscosity and therefore the spreading. Increased spreading was not 
 ob'ained. 
 
 A number of different materials were tested, not only colloids such 
 as saponin, sage tea, gelatin, glue, calcium caseinate, corn, potato, wheat, 
 arrowroot, and starch, but also inorganic substances such as aluminum 
 sulphate with lime, kaolin, barium sulphate with lime, chromium fluoride 
 with lime, calcium chromate with lime, and lead chromate with lime. 
 None of the inorganic materials gave the desired results. On the basis 
 of compatability, efficacy, availability, cost, and ease of preparation, he 
 arranges the colloids in the following order of merit : Calcium caseinate, 
 glue, gelatin, soap bark (saponin), and oil emulsion. 
 
 A new method for testing spreading in the laboratory is presented. 
 Tubes line'd with a coating of wax obtained from the surface of Ben 
 Davis apples were used to determine the capillary rise of the different 
 solutions. The numerical value of 3.5 (probably millimeters) is given 
 for water, while 2 per cent gelatin had a value of 11.8, 2 per cent soap 
 bark 2 8, 2 per cent caseinate 10. 11, and 2 per cent glue 7.8. Lovett be- 
 lieves that the height to which the liquid rises in the capillary tube is a 
 measure of spreading power. This is not strictly true, as only those 
 liquids which will spread over or "wet" the surface of the tube (in this 
 case the wax of the apple) will rise in it. and the height to which they 
 rise is detennined by the surface, tension of the liquid and the angle of 
 contact of the liquid to the wall of the tube. The use of a capillary tube 
 will merely show whether the liquid will or will not spread. If it rises 
 at all, spreading will occur ; and if no spreading will occur there will be 
 no rise. 
 
 Summarizing briefly the work of the various investigators, it is 
 apparent that various colloidal substances, such as casein, gelatin, flour, 
 glue, and soaps modify the spray mixture to such an extent that it 
 will remain spread out over the leaf surface in the form of a film. The 
 surface tension relationships of the leaf and spray offer the best ex- 
 planation, but in certain cases other factors are thought to influence the 
 
SPREADING AND ADHEREXCE OF ARSENICAL SPRAYS 15 
 
 results. These various colloidal substances have given good spreading 
 on leaves with a thin cellulose epidermis, and even on leaves more or 
 less strongly cutinized, such as those of apple, peach, grape, pear, and 
 the citrus fruits. Results on waxy surfaces, such as the waxy bloom on 
 the fruit of apple or plum and the leaves of cabbage, have not been at- 
 tained except by the use of soap or soap emulsions. Soap will produce 
 spreading on such surfaces, but owing to the reaction between the soap 
 and the arsenical materials, its use may result in injury to the foliage. 
 Other disadvantages of the use of soap are its varying composition, its 
 reaction with hard water, and the difficulty of dissolving it without the 
 use of heat. The following experiments were therefore undertaken to 
 find a new material for use on waxy surfaces, and to obtain a better 
 understanding of the causes governing spreading. 
 
 THE THEORY OF SPREADING 
 
 The manner in which surface tension is related to spreading is 
 shown by the following illustration. Upon a liquid, B, with a surface 
 exactly two square centimeters in area, place a drop of an immiscible 
 liquid, A, so that its area is exactly one square centimeter. The area of 
 the surfaces of A, B, and the surface at the interface AB is then 
 exactly one square centimeter in each case, and the surface tension of 
 the entire system is expressed by the equation b A-f- b B-|- b AB, in 
 which b represents the surface tension. Now assume that A is spread 
 out until it completely covers B, then the area of the surface of A and 
 the surface at the interface AB is in each case two square centimeters, 
 while B no longer has an exposed surface. The surface tension of the 
 system then becomes 2 '6 A-\-2 '6 AB. Subtracting the second ec|uation 
 from the first gives the equation b B — b A — b AB as the difference be- 
 tween the surface tensions of the system under the two conditions. 
 Since a system will always arrange itself so as to reduce its surface 
 tension, if the surface tension of the system in the second state is less 
 than in the first case, spreading will occur and the equation will be posi- 
 tive. If the surface tension mirier the original arrangement is the low- 
 est, then the equation is negative and spreading will not occur. The 
 surface tension conditions favorable to spreading may then be repre- 
 sented by the equation b B> b A-|- b AB, while conditions unfavorable 
 to spreading may be expressed by the equation b B< b A-|- b AB. An 
 excellent example of these conditions is the system benzene and water. A 
 drop of pure benzene placed on the surface of pure water immediately 
 spreads out, forming a thin film^. 
 
 bH20> « QH„ + « (H^O-CHr,)-" 
 b=72.o5> « =28.17 -f «=34-68. 
 
 ^ The surface tension values given are those of Harkins, Brown, and Davies (1917). 
 
i6 TECHNICAL BULLETIN 2 
 
 In a few minutes the water becomes saturated with benzene and the 
 benzene becomes saturated with water. The surface tension of each is 
 lowered, while the surface tension at the interface is increased. The 
 value for the equation becoming : 
 
 « Sat. H,0< « sat. CeHo+ b (H3O— QH,). 
 « =6o.i9< « =27.90 4- « =44 .68. 
 The benzene which had previously been spread out in a thin film now 
 collects in the form of a drop. In this case, at least, there is no doubt 
 that spreading or lack of spreading may be explained on the basis of 
 surface tension. 
 
 Unfortunately it is at present impossible to measure the surface ten- 
 sion of a solid, or even to prove conclusively that a solid possesses sur- 
 face tension. Since the surface tension of a liquid increases as the tem- 
 perature is lowered, even to the freezing point, it does not appear rea- 
 sonable that surface tension completely disappears in the solid. It 
 seems more reasonable to believe that a solid has a high surface ten- 
 sion which can not be demonstrated because of the immobility of the 
 molecules. The surface tension at the interface of the solid and liquid 
 is also impossible of measurement. With two of the surface tensions 
 unknown, it is useless to measure the surface tension of the spray 
 liquid, and it becomes necessary to find some other method of solving 
 the problem. 
 
 From a consideration of the above formula, it appears that spread- 
 ing can be obtained only by lowering the surface tension of the liquid or 
 the surface tension at the liquid-solid interface, since the surface ten- 
 sion of the solid (leaf) can not be increased. The problem, therefore, 
 resolves itself into this : How may the surface tension of the spray 
 and the surface tension at the spray-leaf interface be lowered? It is 
 well known (Willows and Hatschek, 191 5) that the adsorption of a 
 solute in the surface layer will result in the lowering of the surface ten- 
 sion, while if it is negatively adsorbed, the surface tension will be in- 
 creased. Thus soap or saponin dissolved in water is adsorbed or col- 
 lects at the surface and the surface tensi^i lowered. Sodium chloride dis- 
 solved in water concentrates in the interior of the liquid and is therefore 
 negatively adsorbed and the surface tension is increased. In a similar 
 manner a dissolved substance adsorbed in the surface at the interface of 
 the liquid and solid will lower the surface tension at this interface. The 
 next question then is to determine what substances will be positively 
 adsorbed in the liquid-air surface and the liquid-solid surface. 
 
 Data on this problem have been given in three different papers which 
 are too long to be completely reviewed here. Altho the papers deal 
 with a different problem, certain facts given in them may be directly 
 applied to the present problem. Harkins, Brown, and Davies (1917). 
 considering the work done when water and some other liquid come 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS i? 
 
 together to form an interface, have shown that at the interface of the 
 two hquids the molecules in the surface of the liquid orient themselves 
 so that the active group is in contact with the water. The decrease in 
 the free energy, i.e., the surface tension at the interface, when the sur- 
 face of a second liquid approaches that of water, depends primarily 
 on the most active group present in the molecules and secondarily on 
 the shape and size of the molecules. The solubility in water is related 
 to this decrease in free energy which more or less perfectly measures the 
 attraction of the active group for the water molecule. The presence of 
 such active groups as COOH, CO, CN, OH, or CONHo is sufficient 
 to make the molecule of an organic substance soluble in water, providing 
 the active group does not have to pull into the solution a slightly less 
 active group, which is too long or too large. 
 
 In a second paper Harkins, Davies, and Clark (1917) show that in 
 the surface of a liquid the molecules are so oriented that the least active 
 group is toward the vapor phase, that is, the vapor above the liquid. At 
 any surface or interface, the orientation of the molecules is such as to 
 make the transition to the adjacent phase — that is, the vapor, liquid, or 
 solid in contact with it — least abrupt. At the interface of two liquids 
 their like parts come together in conformity with the above law. In 
 two liquids, A and B, the portion of the molecules of A like B turn 
 toward B and those of B like A turn toward A. If a solute is present 
 in either of the liquids, A or B, which collects at the interface, the mole- 
 cules of the solute will so arrange themselves that the parts of the mole- 
 cules which are like A will turn toward A, while those like B wall be 
 attracted to B. In such a case the free surface energy, that is, the 
 surface tension at the interface, will be reduced. An aqueous solution 
 of a solute which is active like water but to a lesser degree, will be posi- 
 tively adsorbed in the surface and the least active portion will be turned 
 toward the outside. 
 
 The third paper by Langmuir (1917) approaches the problem from 
 a slightly different angle. Studying the spreading of oleic acid on water, 
 the conclusion is reached that the active carboxyl groups are actually 
 dissolved in the water but the long hydrocarbon chains have too much 
 attraction for each other to be dissolved, hence the acid spreads out into 
 a thin film. Oils without this carboxyl group should not spread, and in 
 fact pure paraffin oil, which lacks an active group, fails to spread at all. 
 Results show that the spreading of films on surfaces is determined by 
 the shape of the molecules and by the relative activities of the different 
 portions of the molecule. In a similar manner a substance, such as a 
 soluble fatty acid, dissolved in water, is adsorbed in the surface layer 
 and lowers the surface tension. Adsorption of a liquid on a solid is 
 considered as similar to the spreading of oil on water, i.e., the portion 
 
i8 TECHNICAL BULLETIN 2 
 
 of the liquid similar to the solid is attracted by the molecules of the 
 solid forming a molecular layer upon it. 
 
 Considering the data presented in these papers, the investigation be- 
 came a search for those substances which possess an active group to 
 render iheni soluble in water and an inactive group which will cause 
 their adsorption in the surface layer. If the inactive group should be 
 similar to the surface of the leaf or soluble in it, then adsorption at the 
 spray-leaf interface would occur and would result in a lowering of the 
 interfacial tension. A substance fulfilling one or the other of these 
 conditions, preferably the latter, should cause spreading. 
 
 SPRAYS THAT WILL SPREAD ON CABBAGE LEAVES 
 
 An eliort was made by the author in the earlier experiments to de- 
 termine whether a given solution would spread or not, by means of cap- 
 illarity. Altho this method might be used in considering waxy sur- 
 faces, it could not be adapted to the study of leaves with a thick cuLicle. 
 To obtain comparable results, nothing but a capillary tube possessing 
 walls exactly similar to the leaf surface would suffice. Tests in all the 
 experiments were therefore made by spraying the solution on the sur- 
 face of the leaf by means of an atomizer. 
 
 The first efforts were directed toward producing spreading of the 
 spray on cabbage leaves. A saturated solution in water of several or- 
 ganic compounds was tested. Solutions of kerosene, methyl salicylate, 
 benzaldehyde, chloroform, camphor, and salicylic acid failed to give the 
 desired results. Amyl alcohol and benzyl alcohol gave good results but 
 were too expensive for actual use. Crude phenol produced spreading but 
 since a i per cent solution of pure phenol failed, the results with crude 
 phenol must have been due to impurities. Phenol possesses an active OH 
 group, while resorcinol possesses two such OH groups. A i per cent 
 solution of resorcinol failed to give spreading. It appeared therefore 
 necessary to introduce such groups in the phenol as would decrease its 
 solubility in water and increase its solubility in the wax of the cabbage 
 leaf, thereby increasing its adsorption at the leaf-spray interface and 
 lowering the interfacial surface tension. A saturated solu'ion of 
 beech wood creosote consisting of a mixture of cresols, C6H4(OH)CH,, 
 and guaiacol, Ce.H^COH lOCHg, gave excellent results. So concen- 
 trated a solution of beechwood creosote is not necessary, since a good 
 spreading accompanied the use of a one-third and even a one-fourth 
 per cent solution. One-fourth per cent is about the lower limit and one- 
 third is a safer proportion. A saturated solution of carvacrol, 
 C6H3(OH(CHo)C3H-. containing about one-twelfth of i per cent, 
 and a saturated solution of eugenol, CeH,(OH) (OCH3)CHo • CH ; 
 CHo. containing onlv about one-fifteenth to one-twentieth of i per 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 19 
 
 cent, gave good spreading, each being less soluble in the water and more 
 soluble in the wax of the leaf.'* 
 
 Most of these substances are adsorbed in the surface layer and re- 
 duce the surface tension of the solution. The question naturally arises 
 whether the spreading is not caused by the lowering of the surface ten- 
 sion of the liquid itself rather than the lowering of the surface tension 
 at the interface of the leaf and the solution. Surface tension determina- 
 tions were therefore made by means of a stalagmometer. Harkins and 
 Brown (1918), in studying this method of determining surface tension, 
 pointed out that if the formation of the drop is too rapid, some of the 
 liquid streaming from the tube seems to force its way into the falling 
 drop during the time of detachment, hence increasing its weight. From 
 the figures given, this error is not very large and can be ignored in 
 these relative determinations. It is also stated that when dilute aqueous 
 solutions of long chain organic molecules are used, it is often necessary 
 for the drop to hang for half an hour to obtain the value of the static 
 surface tension. In spraying it is really the surface tension of the so- 
 lution at the moment it is sprayed over the leaf that is desired, hence a 
 determinatin of dynamic surface tension is more important than static 
 surface tension. 
 
 A one-third per cent solution of creosote allowed to drop at the 
 rate of one drop per minute showed a surface tension of 49.27 dynes. 
 One-fourth of a pound of sodium oleate to 100 gallons of water pos- 
 sessed a surface tension of 42.42 dynes. The soap solution, altho it 
 possessed a lower surface tension than the creosote solution, failed to 
 spread over cabbage leaves. A soap solution with a surface tension of 
 24.28 dynes gave spreading. It is therefore apparent that the spread- 
 ing of the creosote solution is not owing to its low surface tension but 
 to the lowering of the surface tension at the leaf-solution interface. 
 Further confirmation of this statement is found in the fact that none of 
 these solutions would spread over the leaves of citrus, guava, or other 
 plants possessing strongly cutinized leaves but lacking the waxy 
 surfaces. 
 
 The possibility that the results of Lovett (1918) with sage infusions 
 may have been due to some organic principle in the essential oil of the 
 sage, at once suggested itself. Robak (1912) has shown that another 
 species of sage possessed an oil rich in cineol, hence a saturated solu- 
 tion of cineol was tried and gave good results. Since this experiment 
 was conducted, Lovett (1920) has stated that sage infusions did not 
 spread well over the waxy surface of apples, hence his results may 
 have been due to some other ingredient of the infusion. 
 
 * A saturated solution of a mixture of xylenols sold under the name of cresylic acid 
 (straw-colored) by the Barrett Co. will cause spreading on cabbage leaves. 
 
20 TECHNICAL BULLETIN 2 
 
 No doubt a very large number of organic compounds will give simi- 
 lar results. Apparently all that is needed is a chemical containing an 
 active group to make it slightly soluble in the water, and an inactive 
 portion soluble in wax. With this principle as a base, it should be a 
 simple matter to select the compound best suited for each particular set 
 of conditions. The practical application of these results has been left 
 for future investigation. 
 
 SPRAYS THAT WILL SPREAD ON OTHER LEAVES 
 
 It has been stated that these organic solutions failed to spread on 
 leaves which did not possess a waxy coating. Other investigators have 
 obtained good spreading on such leaves by means of various colloids, 
 such as soap, casein, glue, and gelatin. Sodium oleate, 2 pounds to 100 
 gallons, gave spreading on citrus leaves, while a smaller quantity failed. 
 Commercial casein was dissolved in water containing a very small 
 quantity of sodium hydroxide (2 ounces to 5 pounds of casein). Such a 
 solution containing only one-fourth pound of casein to 100 gallons of 
 water gave very good spreading on citrus leaves, but failed on cabbage 
 unless applied with such force that the liquid penetrated between the 
 wax particles, reaching the epidermis below. Gelatin used at the same 
 rate also gave favorable results on citrus leaves. Acidifying the solu- 
 tion did not destroy its spreading properties. 
 
 Flour paste has often been used as a spreader, but much of the 
 protein in the flour is insoluble in water and is therefore not utilized. 
 If about one ounce of sodium hydroxide is added to 5 pounds of flour 
 when the paste is being prepared, all of it goes into solution. Prepared 
 in this way, one-fourth pound of flour to each 100 gallons of the spray 
 produces a film when sprayed over citrus leaves. 
 
 Saponin gives good results but is too expensive for general use. 
 Jacobson (1919) has shown that alfalfa hay contains a saponin quite 
 soluble in water. About one per cent of pure saponin was recovered 
 from air-dried alfalfa. Alfalfa hay should therefore serve as a spreader. 
 One hundred grams of hay was heated in 3 liters of water and allowed 
 to stand twenty-four hours. Two and a half liters were recovered on 
 straining. This solution was diluted until it represented only six- 
 tenths of I per cent of alfalfa hay, and at this dilution gave good 
 spreading. Five pounds of dry alfalfa would be sufficient for 100 gal- 
 lons of the spray. 
 
 The question arose as to whether these results were due to the 
 lowering of the surface tension of the liquid, to the increased surface 
 viscosity suggested by the work of Vermorel and Dantony, or to a con- 
 centration of the materials at the leaf-spray interface, resulting in a 
 lowering of the interfacial tension as suggested by the experiments with 
 waxy leaves. A creosote solution with a surface tension of 49.27 dynes 
 
SPREADING AND ADHERENCE OP ARSENICAL SPRAYS 21 
 
 failed to spread, as did also a soap solution of 38.36 dynes, while one 
 with a surface tension of 24.28 dynes gave good spreading. A solution 
 of gelatin of i part in 10,000 which will produce spreading had a surface 
 tension of 72 dynes. Casein in solution, alkaline to litmus, had a sur- 
 face tension of 64.4 dynes, while when acid to litmus the surface tension 
 was lowered to 57.74 dynes. These values were obtained by allowing 
 one minute for the drop to fall from the tip of the stalagmometer, and 
 therefore represent the dynamic rather than the static surface tension. 
 Since the spreading occurs on the leaf at once, it is apparent that the 
 results are not due to a low surface tension. 
 
 Vermorel and Dantony (1912) explained similar results on the 
 basis of a very rapid surface concentration of the colloid resulting in 
 an increase in the surface viscosity so great that the film was prevented 
 from collecting in a drop. If such a rapid concentration occurred, it 
 should make itself manifest by its influence on the surface tension. 
 The number of drops produced from a given volume of liquid when 
 measured by allowing one second for each drop, compared with the 
 number obtained when one minute was allowed, should be a fair test 
 of the rapidity of the surface concentration which might occur on the 
 leaf surface cjuickly enough to make the surface so viscous that the 
 film could not collect in drops. Water measured at the rate of one 
 drop per second gave 45.5 drops while at the rate of one drop per 
 minute 46 drops were produced. Gelatin, i part in 10,000, gave 46 
 and 46.5 drops; while casein, 3 parts to 10,000, gave 46 and 52 drops 
 when alkaline to litmus and 50 and 58 drops when acid to litmus. This 
 increase does not appear sufficient to produce as high a surface viscosity 
 as Vermorel and Dantony considered possible. Harkins and Brown 
 (1918) fovmd that the surface concentration is so slow in such solutions 
 that a half hour must be allowed each drop to insure the maximum 
 surface concentration and a true measurement of the static surface ten- 
 sion. Further, if the spreading is due to surface viscosity it should be 
 possible to produce the -film on any surface, but this can not be 
 accomplished. 
 
 Like material has been shown to attract like, and there is a strong 
 similarity between the colloids used and the cellulose and cutin of the 
 leaf surface. This similarity should cause a concentration in the sur- 
 face at the interface of the leaf and spray, thus reducing the interfacial 
 tension. On the cabbage leaf the contact is with the wax, different 
 in character from the proteins, and spreading is not obtained. Only 
 when the spray is delivered with sufficient force to drive the liquid 
 between the wax particles and bring it into contact with the epidermis 
 will results be obtained. 
 
22 TECHXICAL BULLETIN 2 
 
 CONCLUSION REGARDING THE PHENOMENON OF 
 SPREADING 
 
 On the basis that substances of hke nature attract each o her, 
 materials were selec'red which, while soluble in water, would be attracted 
 to the leaf surface and concentrate in the leaf-spray interface. Using 
 materials similar in character to the wax of the cabbage leaf produced 
 spreading but failed with leaves which did not possess the waxy coat- 
 ing. Protein and similar materials produced spreading on leaves with 
 a surface of cellulose or cutin but failed to give good sjireading on 
 vvuxy kaves. It appears possible that an infusion of nearly any plant 
 would serve to give spreading to a spray for use on wax-free leaves, 
 but this statement requires confirmation. 
 
 The viscosity of some of the sprays was increased, but in the case 
 of carvacrol, beechwood creosote, and eugenol it was not materially 
 altered. Viscosity, therefore, is not related to spreading. Surface con- 
 centration appears not to take place rapidly enough to increase the 
 surface viscosity to such an extent as to prevent the film from collecting 
 into drops. The surface tension of the leaf was not altered and the 
 values for the surface tension of the liquids did not agree with the 
 results obtained in spreading. The reduction of the interfacial tension 
 appears to be the most reasonable explanation of the facts. Unfor- 
 tunately this surface tension has not yet been measured, nor has the 
 concentration of the material at the leaf -spray interface been dem- 
 onstrated. However, with this theory as a basis, sprays that will 
 spread on the different types of leaves may be produced. 
 
 HISTORICAL CONSIDERATION OF ADHERENCE 
 THE USE OF ORGANIC MATERIALS WITH FUNGICIDES 
 
 The adherence of spray materials was studied very early in the 
 history of spraying. In one of the earliest papers, Millardet and Davis 
 (1886) drew attention to the fact that cupric steatite powder possesses 
 remarkable adherence, owing to its extreme fineness. When it was 
 demonstrated that plant diseases might be controlled by means of 
 copper compounds, an effort was made to form on the leaf a coating 
 of copper which would adhere well and serve as a barrier to the fungus. 
 Crouzet (1899) recommended copper caseinate, prepared by adding 
 raw milk to a solution of copper sulphate, as a very adherent spray. 
 Galloway (1891) used molasses and also glue to give adherence to a 
 mixture of copper sulphate and sodium carbonate. In the experiments 
 of Girard (1892), molasses actually diminished the adherence of 
 bordeaux mixture. Lavergne (1896) tested a "tannocuprique" spray, 
 possibly copper tannate, bordeaux and sugar, and bordeaux mixture 
 
SPREADING AND ADHERENCE OP ARSENICAL SPRAYS 23 
 
 alone, lor the control of black rot. Bordeaux mix. are applied six days 
 earlier than the other sprays gave the most satisfactory results. 
 Cazeneuve (1898) published a paper on albuminous bordeaux mix- 
 tures, the original of which is not available. Bourcart (1913) men- 
 tioned the use by Cazeneuve of three types of albuminous bordeaux, 
 one containing milk, one made with dried egg white, and one widi dried 
 blood. The bordeaux casein did not adhere befter than bordeaux and 
 sugar, but the other two types persisted on the foliage throughout 
 the season. Bourcart cited only the name, hence it is not certain that 
 he is referring to Cazeneuve (1898). Guillon and Guirand (1898) 
 tested the adherence of different preparations to glass slides. They 
 showed that the adherence of bordeaux mixture decreased with its age. 
 The addition of i per cent molasses or three-tenths per cent gelatin 
 increased adherence, while larger quantities decreased it. They also 
 used 3 per cent soap with 2 per cent copper sulphate, a combination 
 which must have resulted in the formation of a colloidal copper soap. 
 In a second paper published the same year they give results of studies 
 on the adherence of these materials to grape leaves. Bordeaux mixture 
 with 2 per cent soap showed the greatest adherence, while gelatin or 
 molasses gave poor results. Perraud (1898) studied the adherence to 
 the fruit rather than to the leaves of the vine. Starch, dextrine, egg 
 powder, and dried blood failed to increase the adherence of bordeaux 
 mixtures. Resin, soap, potassium silicate, molasses, gum tragacanth, 
 and glue increased the adherence in the order named. In a second 
 paper Perraud (1898) recommends the use of resin dissolved in boiling 
 sodium carbonate. This material was added when cool to copper sul- 
 phate to increase its adherence. Similar results were claimed for a 
 mixture of 2 per cent of copper sulphate and 3 per cent of soap. An 
 actual decrease in the adherence of bordeaux mixture due to the addi- 
 tion of turpentine was noted by Ravaz and Bonnet (1903). Resin 
 slightly increased the adherence, but the difference in both cases was 
 considered unimportant. In many of these investigations, the organic 
 material was added to give both adherence and spreading and many 
 workers believed that a spray which would spread would also adhere 
 better. Vermorel and Dantony (1913) clearly distinguished between 
 these two properties, pointing out that a spray may spread without 
 showing any remarkable adherence, while excellent adherence may 
 occur even if the spray does not spread well. Lafforgue (1913) also 
 distinguishes between spreading and adherence. 
 
 These papers represent the development of one method of making 
 funo-icides adhere. In many cases the added material reacted with the 
 copper, forming such compounds as copper saccharate, copper stearate, 
 copper resinate, copper caseinate, and copper albuminate. Some of 
 these compounds, colloidal when freshly prepared, gave an increased 
 
24 TECHNICAL BULLETIN 2 
 
 adherence as a result of the fineness and insolubility of their particles. 
 The copper, being formed into a stable and insoluble compound, lost 
 much if not all of its fungicidal value. This fact coupled with the 
 physiological effects produced by coating the leaves with an insoluble 
 preparation similar to varnish, resulted in the discontinuance to a very 
 large extent of the use of such materials. 
 
 THE USE OF ORGANIC MATERIALS WITH INSECTICIDES 
 
 In early experiments with arsenic preparations for the control of 
 insect pests, investigators feared that the material might adhere too 
 long. Spraying for the gypsy moth developed the use of lead arsenate, 
 a slow-acting poison which could be applied in large quantities without 
 injury to the foliage. Because of its low toxicity, it was necessary 
 that it should adhere for a longer period. Fernald (1894) stated that 
 it is highly desirable to add 2 cjuarts of glucose or molasses to each 150 
 gallons of water to increase the adherence. The idea of using glucose 
 to cause adherence was possibly the result of the use of sugar with 
 fungicides, but lead arsenate gave no reaction with glucose. Lowe 
 (1896) pointed out that the use of glucose did not increase the adher- 
 ence of lead arsenate but that glue at the rate of 2 quarts to 45 gallons 
 gave satisfactory results. Kirkland (1898) showed clearly by means 
 of general observation, chemical analysis, and physiological tests on 
 caterpillars, that the addition of glucose did not increase the adherence 
 of lead arsenate. The glucose was washed ofif the leaves by the first 
 shower. 
 
 Washburn (1891) found that the addition of whale oil or soft soap 
 to the spray increased the adherence of paris green, probably because 
 of its more even distribution. A finely divided solid is more adherent, 
 but Marlatt (1897) showed that paris green when very finely pulverized 
 was more injurious to the foliage. Sirrine (1898) recommended the 
 use of a resin-lime mixture to cause paris green and bordeaux mixture 
 to adhere to the leaves of cabbage and cauliflower. This mixture is 
 really a resinate of lime prepared by making a stock solution of resin, 
 fish oil, and lye, to which, after proper dilution, the lime and paris 
 green are added. The paris green adheres to the particles of the 
 calcium resinate and these in turn adhere to the leaves. Volck (191 3) 
 used flour paste made of 4 pounds of wheat flour to 100 gallons of 
 water to increase the adherence of sulphur. Larger c^uantities of flour 
 caused the dry film to scale off, while smaller quantities were insufificient 
 to "stick" the sulphur. 
 
 Surface (1905) lists the materials used to increase adherence. 
 These are resin-lime mixture, soap, resin added to kerosene and used 
 in kerosene emulsion, molasses, glue, flour, and salt when used with 
 
SPREADING AND ADHERENCE OP ARSENICAL SPRAYS 25 
 
 lime sulphur. The use of salt is now generally recognized as having 
 no influence on the adherence of lime-sulphur sprays. 
 
 Paste lead arsenate was shown by Astruc, Couvergue, and Mahoux 
 (1911) to be more adherent than a freshly prepared suspension of dry 
 lead arsenate. The age of the diluted spray did not appear greatly to 
 influence its adherence. 
 
 Parker (1913) pointed out the possibility of using flour paste with 
 lead arsenate to increase the adherence as well as to produce spreading. 
 A new method of rendering arsenical materials rainproof or more 
 adherent is given in U. S. Patent No. 1 166387. The arsenical substance 
 is coated with an insoluble metallic soap, copper stearate in the case of 
 paris green and lead stearate in the case of lead arsenate. These 
 products were marketed for a few years by the International Color 
 Company, generally as a dry powder for dusting purposes, altho they 
 also manufactured a paste form for general spraying. 
 
 Lees (1915), interested in lime preparations as cover washes for 
 trees in late winter and early spring, tried to increase the adherence by 
 the addition of various substances. Glue, flour, and farina (potato 
 flour) gave good results. The best results were obtained with a mix- 
 ture of whiting, starch, glue, and potassium bichromate. In a second 
 paper (1916) he modified the formula, leaving out the starch. The 
 potassium bichromate was added to the mixture because on exposure 
 to light it reacts with glue, rendering it insoluble. 
 
 O'Kane, Hadley, and O.sgood (1917), in a study of the arsenic 
 retained by fruit after spraying with lead arsenate, give st)me data as 
 to the actual adherence of lead arsenate. Altho not directly concerned 
 with adherence, the percentage of arsenious oxide was determined on 
 fruit picked from three to five days after spraying and before a rain 
 had occurred. The fruit on other trees was not picked until much 
 later, thus giving data for a comparison of adherence. Fruit picked so 
 carefully that none of the lead arsenate was rubbed ofif, showed an 
 average of 0.37 milligram of arsenious oxide per apple, from three to 
 five days after spraying and before rain occurred. An average for 
 twenty-five apples, selected as showing the maximum amount of lead 
 arsenate, was 0.77 milligram per apple. Fruit from other trees sprayed 
 at the same time but picked seventy-five days later had 0.08 milligram 
 and those picked after seventy-six days later had o.ii milligram 
 per apple. These trees were subjected to a total rainfall of 5.01 inches 
 The average amount of arsenious oxide present on the fruit was 25.6 
 per cent. Fruit sheltered by the leaves would no doubt show a much 
 higher percentage of arsenic than would normally be the case. 
 
 Gray (1918) found that the addition of glue to sulphur increased 
 its adherence. The fact that dry lead arsenate was not so adherent 
 as lead arsenate paste and also that the addition of molasses or sugar 
 
26 TECHNICAL BULLETIN 2 
 
 reduced the adherence, was confirmed by the experiments of Hartzell 
 (1918). Stearns (1920) showed that the addition of Irish rock moss 
 or calcium caseinate to the spray increased its adherence. Lovett 
 (1920) considers that in general those substances which increase the 
 spreading of a spray will also increase the adherence of the spray 
 material. 
 
 THE USE OF INORGANIC SUBSTANCES 
 
 The fact that certain inorganic substances adhere to the leaves 
 better than do others^ was early noted. This fact led to the use of 
 certain inorganic materials to increase the adherence of the spray 
 materials. Compounds of chromium, aluminum, and more particularly 
 iron, have been tried. Kilgore (1911) combined london purple and 
 paris green with ferrous sulphate and ferric chloride, but the mixture 
 was not a success. Arsenious oxide combined with ferrous sulphate 
 or ferric chloride gave a rather soluble mixture which proved injurious 
 to the foliage. In these experiments adherence was not sought but 
 rather a combined insecticide and fungicide. Girard (1892) added 
 aluminum sulphate to bordeaux mixture to produce a greater adherence. 
 The use of aluminum sulphate suggested itself to Girard because of 
 its mordant character, but the results were not satisfactory. Fairchild 
 (1894) noted the remarkable adherence of ferrous ferrocyanide. This 
 led to experiments with iron borate, ferric hydrate, and iron sulphide 
 but no such remarkable adherence was noted. Smith (1907) tested 
 an iron arsenate said to be a by-product in the preparation of ore. It 
 contained 45 per cent of arsenious oxide. Used at the rate of T.25 
 pounds to 100 gallons it killed the larvae of the elm leaf beetle without 
 injury to the foliage. The plants were sprayed July 17 and on the 
 i8th a heavy rain occurred and another lighter rain on the 20th. Little 
 trace of the material was found on the 22d. Selby (1908) proposed 
 a modified formula for bordeaux mixture in which the amount of 
 copper sulphate was reduced and a large quantity of iron sulphate was 
 added to serve the purposes of a "sticker." The formula was 
 
 Copper sulphate 2 pounds 
 
 Iron sulphate 4 pounds 
 
 Quicklime 6 pounds 
 
 Water to make 50 gallons. 
 The spray was rust colored and showed great adherence. He 
 considered that the iron sulphate (probably ferrous) was precipitated 
 by the lime in the form of iron hydroxide. Pickering (1907), how- 
 ever, has shown that the reaction of ferrous sulphate and lime forms 
 a precipitate of a basic iron sulphate, evidently of the formula 10 
 FeO.SOs. Volck (1909) prepared an iron sulphide spray by adding 
 ferrous sulphate to lime-sulphur solution. The addition of 20 pounds 
 of iron sulphate to 200 gallons of lime-sulphur spray mixture gave a 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 27 
 
 black muddy precipitate of sulphur, iron sulphide, and calcium sulphate. 
 This spray proved very effective for powdery mildew of apple. 
 
 A definite ferrous arsenate was prepared by Vermorel and Dantony 
 (1909). A solution of 400 grams of ferrous sulphate in 10 liters of 
 water was added and stirred into a solution of 400 grams of sodium 
 arsenate in 10 liters of water, until a test paper impregnated with po- 
 tassium ferro- and ferricyanide became a clear* blue color. Diluting 
 this mixture to 100 liters, it contained 200 grams of ferrous arsenate to 
 the hectoliter. On exposure to the air, as in spraying, the ferrous 
 arsenate forms ferroso ferric arsenate and the ferrous hydrate becomes 
 ferric hydroxide. It is pointed out that ferric hydroxide is an antidote 
 for arsenic poisoning, which fact may explain why very large amounts 
 of this preparation were not injurious to the foliage of plants. Vines 
 were not injured by a proportion of 500 grams to a hectoliter; while 
 pears, apples, and prunes were not injured by a spray containing 2 
 parts in 100. Used at the rate of 100 to 200 grams to i hectoliter, 
 ferrous arsenate successfully controlled the codling moth, Carpocapsa 
 pomoneUa, in experiments extending over three years. Vermorel and 
 Dantony (1909), in another paper, point out that ferrous arsenate can 
 not be successfully combined with bordeaux mixture because of the 
 formation of cupric arsenate, which is injurious to the foliage. They 
 also say that ferrous arsenate alone shows great adherence, but that 
 this adherence is reduced if glucose or molasses is added to the spray. 
 Dalmasso (1910) used ferrous arsenate prepared according to the 
 formula of Vermorel and Dantony for grape spraying in Italy, 
 finding it not quite so effective as lead arsenate. No particular adherent 
 properties are recorded. In America, Smith (1910) used a similar 
 ferrous arsenate, successfully controlling the potato beetle and the elm 
 leaf beetle. No special mention is made of the adherence. About 
 this time several insecticide manufacturers put ferrous arsenate on the 
 market in the United States. One of these, analyzed by Bradley 
 (1910), showed 84.25 per cent of water, only 7.37 per cent of arsenic 
 oxide, and 5.51 per cent of ferrous oxide. The dry ferrous arsenate 
 contains from 50 to 55 per cent of arsenic oxide and from 40 to 45 
 per cent of ferrous oxide. Melander (1911) tested both home-made, 
 and commercial ferrous arsenate for the control of the codling moth. 
 One pound to 125 gallons of water was used, the trees having 0.37 per 
 cent of wormy apples compared with 0.16 per cent for those sprayed 
 with lead arsenate, and 21 per cent for unsprayed trees. Melander 
 said that ferrous arsenate may be successfully combined with lin,e 
 sulphur. Scott and Siegler (1915) tested ferrous arsenate in both 
 laboratory and field experiments. It was found to be a slow-acting 
 poison requiring fairly large doses to insure success. On the other 
 hand, it was considered a safe arsenical substance for use on sensitive 
 
28 TECHNICAL BULLETIN 2 
 
 plants, even 2 pounds of chemically pure ferrous arsenate to 50 gal- 
 lons of water being used on beans and peach without injury. High 
 (1915) considered ferrous arsenate a very desirable spray, since it 
 remains in suspension for long periods and shows no injury on delicate 
 foliage. Ferrous arsenite did not show these desirable qualities. Fer- 
 rous arsenate was successfully combined with the cactus extract used 
 for spreading. 
 
 Howe (1910) tested the iron sulphate-bordeaux mixture recom- 
 mended by Selby (1908), finding it a less effective fungicide than 
 standard bordeaux mixture, but more adherent and less injurious to the 
 foliage. Waite (1910) used two fungicides containing iron sulphate; 
 One a 3-3-50 bordeaux with 2 pounds of iron sulphate, and the other 
 self-boiled lime sulphur plus 3 pounds of iron sulphate. It was 
 observed that iron sulphide spray adhered so strongly as to give the 
 plants a brownish appearance which did not wear off until picking 
 time. It increased the greenness of the foliage and fruit to such an 
 extent that the fruit was late in ripening. Ballard and Volck (1914) 
 pointed out that the chief value of iron sulphide spray as a fungicide 
 is due to the presence of precipitated sulphur, no claim being made of 
 superior adherence. Stewart (1915) mentions the addition of iron 
 sulphate to lime sulphur to reduce the injury to the plants and to 
 increase the "sticking" quality of the spray. Lees (191 5) used alum- 
 inum sulphate, iron sulphate, and potassium bichromate with lime to 
 increase its resistance to rain. Aluminum sulphate was introduced 
 because of the gelatinous character of precipitated aluminum hydroxide. 
 In some cases fair results were obtained but they were not equal to 
 those obtained by the use of organic materials such as glue and flour, 
 
 and hence are not recommended. 
 
 In recent years in a few places in the United States a rain-proof 
 arsenical material has been placed on the market. This preparation 
 contains chromium and arsenic, possibly a chromium arsenate, and when 
 mixed with water forms a bluish-green suspensoid. Great adherence 
 is claimed for this product by the manufacturers. 
 
 Summarizing the review of the literature it is found that in general 
 a finely divided material will adhere better than a coarse material. The 
 use of organic materials to cause spreading of the spray will increase 
 the adherence, apparently owing in part at least to the more even dis- 
 tribution of the material over the leaf surface. Some investigators have 
 found the use of certain inorganic salts or elements, pai'ticularly iron, 
 greatly to increase the adherence of the spray material, while others 
 have failed to obtain these results. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 29 
 
 A NEW THEORY OF ADHERENCE 
 
 A survey of the literature fails to show any definite theory or any 
 attempt to explain why the adherence of one material to the leaf 
 surface is superior to that of another. It appeared necessary as a point 
 of departure to obtain some idea as to the causes of adherence between 
 different substances. One of the best examples of, adherence suggested 
 was that existing between the particles of two oppositely charged 
 colloids. If a colloid exhibiting a negative electric charge is added in 
 the proper proportions to one possessing a positive electric charge, the 
 particles are attracted to each other, forming masses which separate 
 out as a precipitate. An example is the precipitation of a negative 
 arsenious sulphide sol, a colloidal solution, by a positive ferric oxide 
 sol. The colloidal particles of arsenious sulphide adhere strongly 
 to those of ferric oxide. Suspensoids of dye substances also exhibit 
 either positive or negative electric charges, and when two oppositely 
 charged sols are mixed in the proper proportions, a precipitate is formed 
 without any chemical reaction taking place. The Cjuestion arises as 
 to whether solids when not of colloidal dimensions possess such electric 
 charges. Glass ground to colloidal dimensions exhibits a negative elec- 
 tric charge, and if added to a positive sol in the proper proportions will 
 produce a precipitate. A glass beaker should therefore be negative 
 if it possesses an electric charge. A positive sol placed in a clean glass 
 container will have some of its colloidal particles deposi'ied upon the 
 glass and these adhere so thoroly that they can be removed only 
 by acid. Another example is that of filter paper dipped into a positive 
 sol such as methyl violet. It will be observed that altho the water rises 
 on the filter paper the dye particles do not extend above the surface of 
 the liquid. This is due to their precipitation on contact with the negative 
 filter paper. The paper can be placed in running water and the dye 
 particles will not be washed away. On the other hand, if the paper 
 is placed in a negative indigo sol, the dye rises with the water far above 
 the surface of the liquid and is readily removed by washing in running- 
 water. 
 
 King (1917), after a general review of all the theories of dyeing, 
 reached the conclusion that the reaction between the fabric having one 
 electric charge and a dye oppositely charged furnished the best expla- 
 nation of the results obtained in the dye industry. The various nega- 
 tively charged fabrics are easily dyed by positive dyes ; but if a 
 negative dye is to be used, the cloth must be rendered positive by means 
 of a mordant. The mordant is positive and precipitates itself on the 
 negative surface until a new surface with a positive charge is produced. 
 The negative dye will then readily color the cloth. The mordant 
 may be added to the dye in such quantity that the sol becomes positive, 
 when it will give the desired results. 
 
30 
 
 TECHNICAL BULLETIN 2 
 
 If these electric forces are of siifificient magnitude to cause a dye 
 to adhere to a fabric and withstand numerous washings, then if similar 
 ele'^tric charges exist on the leaf and the arsenical materials, they could 
 be u'ilized to increase the adherence of the spray materials. The first 
 question is, do the leaf and the common arsenical materials in use at 
 present possess electric charges, and, if so, are they the same or opposite 
 in character? 
 
 Apparatus Used in Determining the Migration of Particles in the Electric Field 
 A, electrodes; B, zinc sulphate solution; C and D, tap water; E, suspension to be tested. 
 
 ELECTRIC CHARGE EXHIBITED BY COMMON ARSENICAL 
 
 PREPARATIONS 
 
 Methods of testing the charge on the common arsenical preparations 
 found on the market were considered. The first method taken up was 
 the study of the precipitation of sols of known electric charge by suspen- 
 sions of various arsenites and arsenates. This method, however, may 
 give erroneous results, owing to chemical reactions between the various 
 constituents or to the mutual adsorption of particles with like electric 
 charges. The only method considered free from such objections is the 
 determination of the migration of the particles in the electric field. 
 Sols of the common arsenical materials were prepared by reducing the 
 particles to colloidal dimensions by grinding them with water in an 
 agate mortar. A sol to be tested was placed in a U-tube as shown 
 in the drawing. The stop cocks were closed and the upper end of 
 each arm was washed out and filled with tap water. The tube was then 
 clamped into position and connected with tubes of zinc sulphate, con- 
 taining the electrodes, by means of small pieces of glass tubing filled 
 with tap water. The level of the liquids was adjusted by connecting 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 31 
 
 the zinc sulphate containers by means of another piece of tubing contain- 
 ing water. When the level had been established this piece of tubing 
 was removed, the stop cocks were carefully opened, and the current was 
 turned oit. It was found possible to use a 220-volt direct current with- 
 out the introduction of any resistance apparatus. Within half an hour 
 the position of the suspensoid, indicated by its milky or opalescent 
 appearance, was altered by the migration of the' particles toward the 
 oppositely charged electrode. Since all these materials carried negative 
 charges, they migrated toward the positive electrode, the level of the 
 suspensoid being raised in the positive arm of the tube and lowered in 
 the negative arm. The rate of migration was faster on one side because 
 of the pull of gravity on the particles, while the rate was retarded on 
 the other side. In each case the position of the electrodes was changed 
 and the direction of migration reversed. Lead arsenate, paris green, 
 zinc arsenite, tri calcium arsenate, and magnesium arsenate were all 
 found to be negative. The addition of calcium hydroxide did not alter 
 the negative charge of calcium arsenate. Which arsenical substance 
 possesses the strongest charge could not be determined by the experi- 
 ment, since the rate of migration is governed not only by the magnitude 
 of the charge but also by the size of the particles. The rain-proof 
 colloidal preparation containing chromium and arsenic, previously men- 
 tioned, was tested, but no definite migration occurred even in two 
 hours' time. This electric neutrality may be due to the presence of 
 dextrin, which acts as a protective colloid. 
 
 ELECTRIC CHARGE EXHIBITED BY THE LEAF SURFACE 
 
 Having definitely proved by cataphoresis that the common arsenical 
 preparations possess, when wet, an electric charge negative in character, 
 the next question was the demonstration of an electric charge on the 
 leaf surface. Empirically, the leaf should be negative, since filter paper 
 made of cellulose is negative. Cotton, a plant product, is also negative. 
 Stains, which give good results with leaf tissues, either are positive 
 or are prepared with aluminum compounds so that the finished stain is 
 positive. The epidermis of a plant stained with eosin, which is nega- 
 tively charged, readily washes ovtt in water; while Delafield's Haema- 
 toxylin, prepared so that it undoubtedly is positive, is not removed by 
 washing. These statements, however, do not furnish positive proof of 
 the negative character of the leaf. The possibility of determining the 
 charge on the leaf by migration of its particles, after grinding in water, 
 was discarded, because such a test would merely show that the cellulose 
 and proteins of the leaf tissue are negative. 
 
 If gelatin in the form of a sol is placed in the electric field, the 
 particles of gelatin migrate toward the electrode opposite in character 
 to the charoe carried by the gelatin particles. If a concentrated sol 
 
22 TECHNICAL BULLETIN 2 
 
 of gelatin is placed in the bottom of the U-tube and allowed to set to 
 a gel, thus forming a semi-solid barrier, the passage of a current of 
 electricity across the field will no longer be accompanied by a migration 
 of gelatin particles. Under these conditions a flow of water will take 
 place through the gelatin toward the pole with the same sign as the 
 charge carried by the gelatin. The flow of water may be explained 
 by the fact that when a solid takes on a certain electric charge, the 
 water at its surface assumes the opposite sign. This phenomenon, 
 know^n as electric endosmosis, is well known and is utilized in several 
 industrial processes. 
 
 If water could be made to migrate across a leaf surface or through 
 a leaf by endosmosis, it would give positive proof of the presence of an 
 electric charge on the leaf surface. Aftei several unsuccessful attempts 
 the following method gave good results. A bean leaf was bound tightly 
 over one end of a glass tube and the tube partially filled with water. 
 It was then clamped into positicMi with the leaf end immersed in a 
 beaker of water. One electrode was inserted into the tube and the 
 other was placed near the leaf in the beaker. When the current was 
 turned on, the water passed through the leaf, moving in the direction of 
 the negative electrode. Changing the position of the electrodes changes 
 the movement of the water, about half a cubic centimeter passing 
 through the leaf in half an hour. K few minute holes having been 
 pricked in the upper epidermis of a citrus leaf, it was arranged in a 
 similar manner, and the results obtained were even better, since more 
 than a cubic centimeter of water was transferred in half an hour. The 
 better results appeared to be due to the tougher texture of the citrus 
 leaf, which made possible a tighter union with the tube. These experi- 
 ments show that the leaf surface, when wet. exhibits a negative electric 
 charge. 
 
 PREPARATION OF AN ELECTRICALLY POSITIVE 
 ARSENICAL SUBSTANCE 
 
 Having shown that the arsenates and arsenites in use at the present 
 time are electrically negative in character and that the leaf surface shows 
 a similar negative charge, the question is presented as. to how a positive 
 material may be prepared. Most solids ionize in such a manner that 
 they become negative, and of those which are positive and might be 
 utilized in sprays, only three can be considered.^ Sols of aluminum, 
 chromium, and iron oxides show the desired positive charge. Of these 
 aluminum and iron appeared the most promising. Two methods of 
 
 ^ Besides the three mentioned in this paper, positive arsenicals of lead, zinc, calcium, 
 magnesium, copper, and barium have been made and a patent has be^n granted on these 
 products. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 23 
 
 their utilization were suggested: (i ) spraying the leaves wi.h the posi- 
 tive material, followed by the application of the negative arsenical ma- 
 terial ; and (2) the addition of sufficient positive material to the arsen- 
 ical suspension to make the whole mixture positive in character. The 
 first method is impractical, as it would require two sprayings of the 
 plant. Aluminum hydroxide was precipitated by adding ammonium 
 hydroxide to aluminum chloride, care being taken" that not all of the 
 chloride was converted into the hydroxide. A precipitate of aluminum 
 hydroxide is readily peptized by aluminum chloride forming a positive 
 sol. By filtering the precipitate, but not washing it, enough chloride 
 remained to make the material electrically positive. In a preliminary 
 test, a quantity of this aluminum hydroxide was mixed with lead 
 arsenate and sprayed on young citrus trees in the greenhouse. To 
 serve as a check, similar trees were sprayed with lead arsenate alone. 
 After drying, some of the plants from each set were washed by a strong 
 stream of water from a hose for five hours. A macroscopic examina- 
 tion showed practically none of the spray material on the leaves sprayed 
 with the lead arsenate alone, while the leaves of those sprayed with the 
 lead arsenate-aluminum hydroxide mixture were similar in appearance 
 to the check plants which had not been washed. Similar results were 
 obtained by addition of ferric hydroxide to lead arsenate. The substi- 
 tution of geranium plants for citrus trees gave comparable results. 
 
 It would be difficult to make an estimate of the exact quantity neces- 
 sary to give the desired results. The reactions taking place are more 
 easily demonstrated by two colloids possessing opposite electric charges. 
 An illustration would be the reaction of a negative sol of arsenious 
 sulphide with a positive sol of ferric oxide. One particle of the ferric 
 oxide carrying three positive charges would tend to neutralize the 
 charges of three arsenious sulphide particles, since these carry only 
 one negative charge. When a quantity of the ferric sol is added to an 
 arsenious sulphide sol in such proportions that all the negative charges 
 are not neutralized, no precipitation occurs and the mixture is negative. 
 If the number of positive and negative charges is equal or nearly equal, 
 the mixture produces electric neutrality and a precipitate is formed. 
 If enough ferric oxide is added to produce an excess of positive charges, 
 the mixture remains clear and is positive in character. This latter con- 
 dition is the one sought when aluminum hydroxide or ferric hydroxide 
 is added to lead arsenate to increase its adherence. 
 
 Owing to these difficulties, an effort was made to prepare an arsenical 
 substance which would itself be positive. Grimaux (1884) showed 
 that the precipitate formed when ferric chloride is added to sodium ar- 
 senate or sodium arsenite may be peptized, forming a sol which on 
 dialysis gave a basic ferric arsenate, Fe^ (AsO^). Recently, Holmes 
 
34 TECHNICAL BULLETIN 2 
 
 and Rindfusz (1911), Holmes and Arnold (1918), and Holmes and 
 Fall (1919) have studied the preparation of ferric arsenate sols and 
 gels. It is possible to prepare similar preparations of aluminum and 
 chromium. In none of these papers is there any mention of the char- 
 acter of the electric charge carried by the particles of the ferric, alum- 
 inum, or chromium arsenate. Sols of ferric arsenate and arsenite, 
 aluminum arsenate and arsenite, and chromium arsenate prepared in a 
 similar manner were found to be electrically positive. Ferric arsenate 
 precipitated by adding 90 parts of a one-fifth molar solution of sodium 
 arsenate to 100 parts of a one-fifth molar solution of ferric chloride 
 was centrifuged, washed, and the paste dried. This material, ground 
 to colloidal dimensions, showed a migration of the particles toward the 
 negative poles, demonstrating that they carry a positive electric charge. 
 Aluminum arsenate similarly prepared showed a slight migration toward 
 the negative pole. These and several o:her arsenical materials were 
 used in a field test of adherence. 
 
 FIELD TESTS OF ADHERENCE 
 
 The test of the adherence of dift'erent preparations was made on 
 potato plants. The potatoes were in an open field exposed to wind, 
 dew, and rain. During heavy showers, soil was splashed over the plants 
 and washed off again, actually rending to scour the leaves. In the 
 cultivation of the plants, soil was frequently thrown over the leaves, 
 thus tending to remove the spray materials. 
 
 The following preparations were used in the proportions given, the 
 amount of each material being figured on a dry basis. 
 
 TABLE I 
 
 Pounds to Gallons 
 
 Lead arsenate powder 
 
 Calcium arsenate powder 
 
 Zinc arsenite powder 
 
 Magnesium arsenate powder 
 
 Paris green powder 
 
 Lead arsenate powder, 1.5 lbs. + ferric hydroxide paste, 0.4 lb. (dry 
 
 weight) 50 
 
 Ferric arsenate paste 2.0 50 
 
 Ferric arsenite paste 2.0 50 
 
 Ferrous arsenate paste 2.0 ^o 
 
 Aluminum arsenate paste i .0 50 
 
 The materials were applied to one or two rows of potatoes and 
 several rows throughout the plot were left unsprayed to serve as checks. 
 When the sprayed plants were dry, ten representative leaves were col- 
 lected from each set to be analyzed for arsenic. There was no rain for 
 six days after the sprays were applied and then only 0.06 inch fell, 
 hence all the compounds had an excellent opportunity to destroy the 
 potato beetle larvae. Nine days after spraying the check plants were 
 completely stripped of leaves and a count of larvae on the sprayed 
 plants gave the results shown in Table II. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 35 
 
 TABLE II 
 
 Lead arsenate 
 
 Calcium arsenite 
 
 Zinc arsenite 
 
 Magnesium arsenate . .• 
 
 Paris green 
 
 Lead arsenate + ferric hydroxide. 
 
 Ferric arsenate 
 
 Ferric arsenite 
 
 Ferrous arsenate 
 
 Aluminum arsenate 
 
 Average 
 per plant 
 
 Some of these averages are high, owing to the proximity of check 
 rows where the larvae, having stripped the plan':s, were migrating to 
 adjacent rows. A count on part of the lead arsenate row next to a 
 check row showed an average of four larvae per plant, while that part 
 not adjoining the check row had an average of one larva per plant. 
 This condition existed in the rows sprayed with ferric arsenate, ferrous 
 arsenate, and aluminum arsenate, but not in the row treated with lead 
 arsenate-ferric hydroxide mixture. 
 
 In the preparation of the lead arsenate-ferric hydroxide mixture 
 there was a decided tendency for the particles to form masses which 
 were extremely difificult to hold in suspension. The first six plants in 
 the row obtained a very heavy dose of the spray materials while the 
 rest of the row received very little. For this reason in the analysis of 
 the plants, the first six plants in this set were kept separate from the 
 others. In the analysis of the plant material for arsenic, the organic 
 material was oxidized by boiling with nitric and sulphuric acids. The 
 arsenic was determined by the Gutzeit method (Scott, 1917). The 
 leaves of the plants, taken immediately after the spray had dried, repre- 
 sented 100 per cent of the spray material. Other determinations were 
 made on leaves collected seven, eleven, and seventeen days after spray- 
 ing, the average amount per leaf in terms of arsenious oxide being 
 recorded in Table III. In this table is also given a record of the 
 rainfall during this period. 
 
 Reviewing the field test, a relationship between the adherence and 
 the electric charge is cjuite apparent. Lead arsenate, which adhered 
 best of those possessing a negative electric charge, showed only 6.6 per 
 cent at the close of the experiment. Aluminum arsenate tested in the 
 electric field was only very slightly positive, since only a slight migra- 
 tion of the particles occurred. Ferrous arsenate was negative, but on 
 the plant decomposition would produce ferric arsenate and ferric 
 hydroxide according to Vermorel and Dantony (1909). The decom- 
 posed material would then no doubt be positive. No definite data are 
 available concerning the charge possessed by particles of the lead 
 arsenate-ferric hydroxide mixture, but it was mixed in proportions 
 
36 
 
 TECHNICAL BULLETIN 2 
 
 which should form a positive material. Ferric arsenate was considered 
 the most promising of the entire set, but its use requires further 
 investigation. 
 
 TABLE III 
 
 Lead arsenate 
 
 Calcium arsenate 
 
 Zinc arsenite 
 
 Magnesium arsenate. . . 
 
 Paris green 
 
 Lead arsenate + fer- 
 ric hydroxide 
 
 First 6 plants. . . . 
 
 Other plants 
 
 Ferric arsenate 
 
 Ferric arsenite 
 
 Ferrous arsenate 
 
 Aluminum arsenate... 
 
 Rainfall 
 
 
 
 
 Per 
 
 
 
 Per 
 
 
 
 Per 
 
 After 
 
 
 
 cent 
 
 
 
 cent 
 
 
 
 cent 
 
 spray- 
 
 6 
 
 7 
 
 of 
 
 10 
 
 1 1 
 
 of 
 
 16 
 
 17 
 
 of 
 
 ing 
 
 days 
 
 days 
 
 total 
 
 days 
 
 days 
 
 total 
 
 days 
 
 days 
 
 total 
 
 Mg. 
 
 
 Mg. 
 
 
 
 Mg. 
 
 
 
 Mg. 
 
 
 0.75 
 
 
 O.IO 
 
 40.0 
 
 
 0.200 
 
 26.6 
 
 
 0.050 
 
 6.6 
 
 1. 00 
 
 
 O.IO 
 
 10. 
 
 
 0.050 
 
 5-0 
 
 
 0.050 
 
 5-0 
 
 1.40 
 
 
 0.35 
 
 25.0 
 
 
 0.I2S 
 
 8.9 
 
 
 0.075 
 
 5-35 
 
 1-25 
 
 
 O.I 5 
 
 12.0 
 
 
 0.075 
 
 6.0 
 
 
 0.025 
 
 2.0 
 
 1.60 
 
 
 O.IO 
 
 6.2 
 
 
 0.07s 
 
 4.6 
 
 
 0.025 
 
 1.5 
 
 8.00 
 
 
 2.40 
 
 30.0 
 
 
 
 
 
 1.400 
 
 17-5 
 
 0.65 
 
 
 0.20 
 
 30.7 
 
 
 O.IOO 
 
 15-3 
 
 
 O.IOO 
 
 15-3 
 
 1.20 
 
 
 0.60 
 
 50.0 
 
 
 0.400 
 
 33-3 
 
 
 0.200 
 
 16.6 
 
 2.40 
 
 
 0.75 
 
 31.2 
 
 
 0.600 
 
 25.0 
 
 
 0.300 
 
 12.5 
 
 2.80 
 
 
 1-25 
 
 44.6 
 
 
 0.750 
 
 26.7 
 
 
 0.350 
 
 12.5 
 
 0.75 
 
 
 0.1s 
 
 20.0 
 
 
 0.075 
 
 10. 
 
 
 0.075 
 
 10. 
 
 In. 
 
 0.06 
 
 In. 
 
 0.03 
 
 
 In. 
 
 0.80 
 
 .... 
 
 
 In. 
 2.18* 
 
 
 
 Total 
 rain- 
 fall 
 
 In. 
 307 
 
 * This record represents one severe storm. 
 
 The poor adherence of calcium arsenate is of particular interest. 
 Graham (1918) found it to adhere better than lead arsenate, while 
 Wilson (1919) also considered it equal in adherence to lead arsenate. 
 Graham's tests were conducted in the greenhouse and hard water was 
 used to wash off the arsenical materials. Owing to the presence of the 
 common calcium ion, the presence of calcium carbonate in the tap water 
 would reduce the solubility and the rate of decomposition of the calcium 
 arsenate. Wilson applied the calcium arsenate with lime and a similar 
 effect was obtained. In these experiments, the calcium arsenate was 
 used alone, and apparently rapidly dissolved, decomposed, and disap- 
 peared from the leaves. 
 
 PREPARATION OF POSITIVELY CHARGED FERRIC 
 
 ARSENATE 
 
 In further investigations with ferric arsenate it was found that it 
 did not always possess a positive electric charge, in fact it was often 
 neutral or even negative. An effort was made to discover under what 
 conditions the best material could be prepared. Grimaux (1894) 
 considered the first precipitate formed by the addition of sodium 
 arsenate to ferric chloride as represented by the formula (As04H)3Fe2. 
 When this material was peptized by ferric chloride, a reaction was 
 obtained represented by the following equation : 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS Z7 
 
 2(AsO,H)3Fe2+FeXl6=3 (AsO.H), Fe^Cl^ 
 
 Upon dialyzing this sol a basic compound of ferric arsenate was 
 retained by the membrane while hydrochloric acid passed through. 
 Holmes and Arnold (1918) noted that the appearance of the gel 
 obtained on dialysis depended upon the proportion of ferric chloride 
 used in the peptization of the ferric arsenate precipitate. If a large quan- 
 tity was used the gel obtained was reddish or redxlish-brown, while if 
 only a small quantity was used the gel was a dirty yellowish color. A 
 study of these materials and the character of the charge carried by 
 their particles should throw light upon the problem. 
 
 Three different sols were prepared, the first, which will be referred 
 to as the 50-45, was made by adding slowly to 50 cubic centimeters 
 of a one-fifth molar solution of ferric chloride, 45 cubic centimeters 
 of a one-fifth molar solution of disodium arsenate, the precipitate being 
 peptized as rapidly as it formed. The second was similarly prepared, 
 but only 35 cubic centimeters of the arsenate was added, while 25 cubic 
 centimeters was added to the third. These were known as the 50-35 
 and the 50-25 sols. The 50-45 sol was light yellow with greenish tints 
 and somewhat clouded. The 50-35 and 50-25 sols were perfectly clear 
 and of a deep yellow- color with greenish tints. The three different 
 sols were dialized in separate containers by means of collodion sacs. 
 Only distilled water was used and all the wash water was saved for 
 analysis. Having determined the amount of CI, AsO^, and Fe present 
 in the original sols, a fairly accurate knowledge of the composition of 
 the gel was obtained by the analysis of the wash water. 
 
 Hydrogen sulphide was first passed through the solution to 
 remove the arsenic. This required considerable time to insure its com- 
 plete removal. The arsenious sulphide was filtered off, redissolved, 
 and the quantity of arsenic finally determined by iodine titration. The 
 filtrate was made up to 1,000 cubic centimeters and a portion used for 
 the determination of the ferrous iron by titration with potassium 
 permanganate. The chlorine was titrated with silver nitrate, using 
 potassium chromate as an indicator. Considerable difficulty was ex- 
 perienced with this method since the indicator did not indicate the end 
 point, more chlorine being found than had existed in the original sol. 
 A double titration method using an excess of silver nitrate and then 
 determining this excess by titration with ammonium thiocyanate using 
 ferric sulphate as an indicator gave better results. The three different 
 sols may be represented by the following equations based upon the 
 results of the analyses. 
 
38 TECHNICAL BULLETIN 2 
 
 50-25 Sol 
 
 CI] 
 50 FeCl3+25 Na.H(As04-+XHoO = Fe-Fe(As04) -^Cl [. Fe-Fe(AsOi) +25 HC1+ soXaCl 
 
 CI J 
 Fe. = 0.60256 g. • Fe=o.i344 g. 
 
 CI. = 1.0476 g. As04=o.i53o g. Fe=o. 46816 g. 
 
 AsOi = o.70o6 g. (Colloid) As04 = o.S476 g. 
 
 Cl=i.04i5 g. 
 
 (Passed through membrane 
 crystalloid) 
 50-35 Sol 
 
 cn 
 
 :;o FeCl3-f35 NaoH(As04+XH20 = Fe-X FeAsOi +C1 }. Fe-FeAs04+35 HCl+70 NaCl 
 
 CI I 
 Fe=o. 60256 g. Fe=:o. 22624 g- Fe=o. 37632 g. 
 
 C! = i.0476 g. AsOi^o. 50064 g. AsOi=o.48o2 g. 
 
 As04=o. 98084 g. (Colloid) Cl= 1.0437 g- 
 
 (Crystalloid) 
 50-45 Sol 
 
 CI] 
 so FeCl-!=45 Na2H(As04)+XH20 = Fe-XFeAsOi +C1 |. Fe-FeAs04+45 HCI + 90 NaCl 
 
 CI J 
 Fe = o. 60256 g. Fe = o. 56566 g. Fe:=o.0369 g. 
 
 Cl = i.0476 g. As04=i. 20703 g. As04=o.oS405 g. 
 
 As04=i.26io8 g. (Colloid) Cl=i. 03143 g. 
 
 (Crystalloid) 
 
 From these equations it appears that there is a compound of ferric 
 arsenate and ferric chloride formed which is capable of passing through 
 a dialyzing membrane.'' The proportions in all three cases show that 
 this compound is made up of one molecule of ferric chloride and one 
 of ferric arsenate, ClgFe-Fe (ASO4 ). The proportions of ferric arsen- 
 ate of iron in the colloid, however, vary. In the 50-25 it apparently may 
 be represented by the formula Fe-FeAsO^, while in the other gels a 
 larger quantity of ferric arsenate is present and the formula may 'be 
 represented by Fe X Fe (AsO^),. The 50-25 formula appears to be 
 ferric oxide with an adsorbed molecule of ferric arsenate. The 50-35 
 and 50-45 may contain ferric oxide with an X number of adsorbed 
 ferric arsenate molecules or it may consist of Fe-FeAsO^ with a num- 
 ber of unadsorbed ferric arsenate molecules. The 50-45 sol set to a 
 gel when dialyzed from ten to fifteen hours, and was a rather opaque 
 dirty yellow mass. The 50-35 and 50-25, after two to three days dialy- 
 sis, finally formed a clear reddish gel. 
 
 The three gels were carefully dried, and after grinding to colloidal 
 dimensions their migration in the electric field was determined. Both 
 the 50-45 and the 50-35 migrated toward the positive pole and therefore 
 possessed a negative charge. The 50-25 sol. being positive, migrated 
 toward the negative pole. All three .sols before dialyzing had particles 
 carrying positive charges. 
 
 <■ Arsenic acid may be liberated by the stronger hydrocloric acid, and the material which 
 passes through the membrane may consist of a mixture of ferric cloride and arsenic acid. If 
 such is the case, it is not apparent why this arsenic acid should be reduced in quantity by the 
 addition of sodium arsenate. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 39 
 
 From these results it would appear that the pure or nearly pure 
 ferric arsenate was negative, while if ferric chloride or ferric oxide is 
 present with it, the particles possess a positive charge. Ferric arsenate 
 prepared from equal molecular portions of disodium arsenate and ferric 
 chloride was filtered, washed, dried, and prepared as a suspensoid by 
 grinding. This material was negative. The addition of ferric chloride 
 equal in weight to the ferric arsenate present in the» suspensoid, caused 
 it to take on a strong positive charge. Holmes and Fall (1919) have 
 shown that the amount of ferric chloride necessary to peptize a given 
 quantity of ferric arsenate increases rapidly with the age of the pre- 
 cipitate up to one or two days, with only a smaller further increase 
 being noted with older precipitates. This was found to be due to a 
 decrease in the hydration of the precipitate and in the formation of 
 larger aggregates with a consequent decrease in the external surface of 
 the particles. In the present investigation it was found to be impos- 
 sible to peptize with ferric chloride a precipitate which had been thoroly 
 dried. However, as already shown, the ferric chloride is able to give 
 the particles a positive charge, due apparently to a reaction with the 
 surface of the particles. The amount necessary to give the desired 
 results will vary with the surface area of the particles. The quantity 
 necessary to produce a positive charge on the surface of large particles 
 such as are present in the spray suspension, will be much less than is 
 required where the surfaces are greatly increased by reducing the size 
 of the particles to colloidal dimensions. The minimum quantity that 
 may be used to give the desired adherence to a ferric arsenate spray, 
 remains to be determined by further tests in later investigations. 
 
 Holmes and Rindfusz (1916) showed that ferric arsenate may be 
 peptized by ammonium and other hydroxides, apparently a ferric oxide 
 with an adsorbed arsenate being produced. This method has been tried 
 and gives a positive material but is undesirable, since the presence of 
 large quantities of ferric oxide reduces the toxicity of the mixture. 
 
 COMPARATIVE TOXICITY OF DIFFERENT ARSENICAL 
 PREPARATIONS 
 
 Freslily precipitated ferric hydroxide is a well-known antidote for 
 arsenical poisoning in higher animals (Sollmann, 1917). The reaction 
 between ferric hydroxide and arsenious oxide was first thought to be 
 due to the formation of a basic ferric arsenate, but Biltz (1904) has 
 shown that the ferric hydroxide adsorbs the arsenious oxide and does 
 not form a definite compound. Vermorel and Dantony (1909) stated 
 that ferrous arsenate on the plant was decomposed with the formation 
 of ferric hydroxide. Scott and Siegler (191 5) found ferrous arsenate 
 low in toxicity. In view of these facts it was considered advisable to 
 
40 TECHNICAL BULLETIN 2 
 
 obtain some idea of the toxicity of these new preparations in com- 
 parison with other common arsenical compounds. 
 
 METHODS OF DETERMINING COMPARATIVE TOXICITY 
 
 A short review of the hterature deahng with toxicity presents sev- 
 eral methods of obtaining a comparison of toxicity. Marlatt (1897) 
 studied the toxicity of paris green, london purple, and copper arsenate, 
 using as a basis of comparison the time required to kill the insects and 
 the total number killed. Holloway (1912) attempted to express th.^ 
 toxicity numerically. Paris green at the rate of two milligrams per 
 leaf was taken as a standard. The toxic value was expressed in the 
 period of time required to kill. If a given poison required twice as 
 long as paris green to produce death, then its toxic value would be 
 expressed as 0.5, while a poison killing more quickly than the standard 
 would be represented by a number greater than i. This value was 
 called the poison exponent. Mention is made of iron arsenate having 
 a poison exponent for Hcliothus obsolcta of 0.5 for the first instar, 
 0.33+ for the second, 0.81 for the third, and 0.47+ for the fourth, fifth, 
 and sixth instars. Scott and Siegler (191 5) compared the killing 
 values of various arsenical compounds by measuring in square inches 
 the amount of leaf surface consumed by the fall webworm, Hyphantria 
 cunea (Drury), feeding on black cherry, Pruniis serotina. The time 
 required to produce death was also taken into consideration. They 
 consider lead arsenate as the best arsenical compound, but the triplumbic 
 form is a slow acting poison. No attempt is made to arrange the prep- 
 arations in the order of their toxicity and such an arrangement would 
 be difficult, owing to the variation in quantity of the poisons used. 
 Tartar and Wilson (191 5) studied the comparative toxicity of lead 
 hydrogen arsenate and basic lead arsenate. An arsenic analysis was 
 made of the tissues of the dead insects, showing that those killed by 
 the acid lead arsenate contained more arsenic. This was considered 
 due to the more rapid adsorption of the chemically reactive diplumbic 
 form, while the triplumbic lead arsenate was passed through the intes- 
 tinal tract without much absorption taking place. Sanders and Brit- 
 tain (1916) based their comparison of toxicity upon the percentage of 
 larvae dead after feeding on sprayed foliage for a certain arbitrary 
 number of days. Lovett and Robinson (1907), continuing the investi- 
 gations started by Lovett and Wilson (1915), take into consideration 
 the time required to kill, the approximate amount of the arsenical com- 
 pound necessary to produce death, and the ratio of arsenic pentoxide 
 in the tissues to that in the excrement. The ratio of lead hydrogen 
 arsenate is given as i to 0.544, calcium arsenate i to 0.70, and basic 
 lead arsenate i to 1.51. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 41 
 
 COMPARATIVE TESTS OF TOXICITY 
 In the present work, locusts, Mdanoplus fcmur-ruhnun, were fed 
 upon a standard mixture composed of i gram of bran to 0.5 gram of 
 sugar, 1.5 cubic centimeters of water, and 0.04 gram of the poison to be 
 tested. When the mixture became dry it was moistened with a few 
 drops of water. Large glass cylinders covered with cheese-cloth at 
 the top were used as feeding cages. Each cage contained ten insects. 
 Numerous observations were made, and the number of dead locusts 
 was recorded. When all were dead in any experiment their bodies and 
 their excreta were separately collected, dried, and preserved for analysis. 
 The experiments with each poison were repeated five times, hence the fig- 
 ures given represent the average for fifty locusts. The relative merit of 
 the different methods of comparing toxicity may be judged from the 
 data presented, which are based upon the time required to kill, the 
 amount of bran mash consumed, and the ratio of the poison found in 
 the body and in the excreta. The organic matter was destroyed by 
 boiling with nitric and sulphuric acid and the arsenic determined by the 
 Gutzeit method. (Scott 1917). The data have been compiled in the 
 following tables, where the materials are arranged in order of their 
 toxicity, the most toxic substance being placed first. 
 
 These tables show clearly that the best basis of comparison is the 
 ratio of the quantity of the chemical found in the body to that found 
 in the excreta. This point is confirmed by feeding mixtures containing 
 different quantities of the same chemical. 
 
 TABLE IV 
 Toxicity Based on Number of Grams of Bran Mash Consumed Before Death* 
 
 Calcium arsenate 0.0045060 
 
 Sodium arsenite 0.0051975 
 
 Calcium arsenite 0.0052730 
 
 Paris green 0.0056700 
 
 Ferric arsenite 0.0058905 
 
 Colloidal aluminum arsenate 0.0060705 
 
 Zinc arsenite 0.0061905 
 
 Colloidal ferric arsenate 0.0063370 
 
 Magnesium arsenate 0.0077400 
 
 Stearated paris green 0.0077400 
 
 Ferric arsenate 0.0078000 
 
 Sodium arsenate 0.0078000 
 
 Aluminum arsenate 0.0083610 
 
 Arsenious oxide 0.0116700 
 
 Ferrous arsenate 0.0120000 
 
 Lead arsenate 0.0210000 
 
 * Calculated from the sum of the arsenic recovered from the bodies and excreta and 
 the proportion of arsenic present in the mash. 
 
TECHNICAL BULLETIN 2 
 
 TABLE V 
 
 Toxicity Based on the Average Number of Hours Necessary to Produce Death 
 
 Sodium arsenite 20.1 
 
 Calcium arsenite 25.5 
 
 Stearated paris green ,^0.3 
 
 Arscnious oxide 33.0 
 
 Sodium arsenate 37.5 
 
 Paris green 38. i 
 
 Magnesium arsenate 44-7 
 
 Calcium arsenate 45-^ 
 
 Ferric arsenate 46. 7 
 
 Zinc arsenite 46.7 
 
 Colloidal aluminum arsenate 57. 2 
 
 Colloidal ferric arsenate 57.9 
 
 Ferric arsenite 60.6 
 
 Aluminum arsenate 67.2 
 
 Ferrous arsenate 70.0 
 
 Lead arsenate 92.2 
 
 TABLE VI 
 Toxicity Based on the Ratio of the Chemical in the Body and in the Excreta 
 
 Chemical 
 
 Arsenious oxide 
 
 Sodium arsenite 
 
 Colloidal aluminum arsenate. 
 
 Sodium arsenate 
 
 Magnesium arsenate 
 
 Calcium arsenate 
 
 Colloidal ferric arsenate 
 
 Calcium arsenite 
 
 Stearated paris green 
 
 Ferric arsenite 
 
 Zinc arsenite 
 
 Paris green 
 
 Ferric arsenate 
 
 Aluminum arsenate 
 
 Ferrous arsenate. . '. 
 
 Lead arsenate 
 
 Milligrams 
 in excreta 
 
 0.005600 
 0.004330 
 0.006051 
 
 O.OIOOOO 
 
 0.006400 
 0.005290 
 o.oi 1440 
 0.006100 
 0.010900 
 0.010578 
 0.017800 
 0.015000 
 0.048000 
 0.047160 
 
 0.120000 
 
 0.486660 
 
 Milligrams 
 in body 
 
 0.15000 
 o.i IIIO 
 0.15130 
 0.25000 
 0.16000 
 0.13200 
 0.21440 
 o. II 000 
 o. 16000 
 0.13210 
 0.1 32 10 
 
 O.IIOOO 
 
 0.16000 
 0.13830 
 0.20000 
 0.44333 
 
 Excreta 
 
 Ratio 
 
 Body 
 
 0.037334 
 0.039000 
 0.039995 
 0.040000 
 0.040000 
 0.040077 
 0.053360 
 0.055455 
 0.068130 
 0.080077 
 0.134750 
 0.136370 
 0.300000 
 0.341240 
 0.600000 
 1.097700 
 
 TABLE VII 
 
 Grams of lead 
 
 Grams of bran 
 
 
 Excreta 
 
 arsenate added to 
 
 consumed before 
 
 Hours required 
 
 T> *-'., 
 
 
 bran mash 
 
 death 
 
 kill 
 
 Body 
 
 0.02 
 
 0.0300 
 
 66.5 
 
 I.OOCO 
 
 0.04 
 
 0.0210 
 
 92.2 
 
 1.0998 
 
 O.IO 
 
 0.0108 
 
 69.5 
 
 1.2005 
 
 0.20 
 
 0.0090 
 
 55.5 
 
 1.2502 
 
 0.30 
 
 0.0070 
 
 58.0 
 
 0.9092 
 
 The hour.s required to produce death vary greatly while the quan- 
 tity consumed is reduced rapidly. On the other hand, the results are 
 affected but slightly by great differences in the quantity of the poison 
 fed to the locusts. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 43 
 
 A METHOD OF EXPRESSING TOXICITY 
 Incidentally this method opens up a possibility of directly comparing 
 arsenical compounds, giving them a definite numerical value in com- 
 parison with some standard compound of arsenic. In the present 
 investigation, the object was merely to obtain an idea of the relative 
 toxicity of certain new arsenical materials compared with those in use 
 at the present time. By taking the reciprocal of the number obtained 
 by dividing the quantity of the chemical in the excreta by that found 
 in the body, the most toxic chemical becomes the one with the highest 
 numerical value, as is shown in Table VIII. 
 
 TABLE VIII 
 
 Arsenious oxide , 
 
 Sodium arsenite 
 
 Colloidal aluminum arsenate 
 
 Sodium arsenate 
 
 Magnesium arsenate 
 
 Calcium arsenate 
 
 Colloidal ferric arsenate... 
 
 Calcium arsenite 
 
 Stearated paris green 
 
 Ferric arsenite 
 
 Zinc arsenite 
 
 Paris green 
 
 Ferric arsenate 
 
 Aluminum arsenate 
 
 Ferrous arsenate 
 
 Lead arsenate 
 
 Reciprocal 
 
 of the ratio 
 
 Value when 
 lead arsenates 
 
 26.78500 
 
 25.64100 
 
 25.00000 
 
 25.00000 
 
 25.00000 
 
 24.95200 
 
 18.74000 
 
 18.03200 
 
 14.78000 
 
 12.48800 
 
 7.42110 
 
 7-33330 
 
 3-33330 
 
 2.93050 
 
 1.66660 
 
 0.91099 
 
 29.4010 
 
 28.1460 
 
 27.4430 
 
 27.4430 
 
 27.4430 
 
 27-3890 
 
 20.5710 
 
 19.7940 
 
 16.1120 
 
 13.7070 
 
 8.1462 
 
 8.0494 
 
 3-6590 
 
 3.2169 
 
 1.8295 
 
 1. 0000 
 
 Assuming the value of the lead arsenate to be i in.stead of 0.91099 
 it is simple to express the toxicity of any of the other arsenical com- 
 pounds in terms of lead arsenate. Ferric arsenate would then have a 
 value of 3.659, or in other words, would be more than three times as 
 toxic as lead arsenate, while paris green would be eight times, and 
 arsenious oxide more than twenty-nine times as toxic as lead arsenate. 
 Selecting absolutely pure lead arsenate as a standard, several common 
 insects could be fed under definite conditions and a numerical value 
 applied to any compound of arsenic placed on the market. 
 
 The values presented in this paper can not be used in this manner, 
 as the lead arsenate used was a commercial product and not chemically 
 pure. 
 
 INFLUENCE OF FERRIC HYDROXIDE ON TOXICITY 
 
 The results obtained by the addition of ferric hydroxide to arsenious 
 oxide are recorded in Table IX. 
 
44 TECHNICAL BULLETIN 2 
 
 TABLE IX 
 
 Ratio 
 Chem. in excreta 
 
 Chem. in body 
 
 Arsenious oxide 0.04 gram 0.037334 
 
 Arseniovis oxide 0.04 g. + stale ferric hydroxide paste containing 0.1614 g. 
 
 dry weight 0.056000 
 
 Arsenious oxide 0.04 g. + colloidal ferric oxide containing 0.0647 g. of 
 
 dry material 0.133333 
 
 Arsenious oxide 0.04 g. + fresh ferric hydroxide paste containing 0.1714 
 
 g. dry weight 0.333333 
 
 Arsenious oxide 0.04 g. ± fresh ferric hydroxide paste containing 0.17 14 g. 
 
 dry weight, but allowed to stand mixed for 48 hours before feeding.... 0.600060 
 
 The data show clearly that the addition of ferric hydroxide which 
 adsorbs arsenious oxide reduces the toxicity of the mixture. Calcium 
 hydroxide mixed in the same way gave a ratio of 0.05 compared with 
 calcium arsenite with a ratio of 0.055455. In this case the two reacted, 
 giving calcium arsenite with a ratio checking with the ratio of calcium 
 arsenite previously tested. None of the figures given in the table 
 check with ferric arsenite having a ratio of 0.080077. This compound 
 is apparently not formed. 
 
 The addition of ferric hydroxide to other arsenical compounds gave 
 similar results. Sodium arsenate with a ratio of 0.039 ^^'^s reduced 
 by the addition of ferric hydroxide to 0.1333, while paris green 0.13637 
 became 0.74834. 
 
 These experiments explain the cause of the low toxicity of ferrous 
 arsenate. Ferric arsenate is not so low in toxicity, but the addition of 
 ferric hydroxide in any large quantity to increase the positive electric 
 charge would no doubt reduce its toxicity. The adsorption of arsenic 
 preparations by ferric hydroxide may be the explanation of the lack 
 of injury to bean and peach foliage recorded by Scott and Siegler 
 (1915). It has not been possible to investigate this phase of the 
 subject at the present time. 
 
 CONCLUSION REGARDING THE PHENOMENON OF 
 ADHERENCE 
 
 In that portion of the present investigation dealing with the adher- 
 ence of spray materials to the leaf, it has been demonstrated that the 
 leaf surface assumes, when wet, a negative electric charge and that 
 suspensoids of the common arsenic compounds ionize in such a way 
 that their particles are also negative. Based upon the results obtained 
 in the dye industry, the assumption is made that spray materials carry- 
 ing positive electric charges would adhere to the negatively charged 
 leaf surface better than materials exhibiting negative charges. Field 
 tests have confinned this assumption. Positive arsenic preparations 
 
SPREADIXG AXD ADHEREXCE OF ARSENICAL SPRAYS 45 
 
 of different elements were prepared and tested, ferric arsenate being 
 considered the most promising material. The presence of ferric 
 hydroxide in the spray material is not desirable since, owing to its 
 adsorption of arsenic, it lowers the toxicity of the preparation. 
 
 Many phases of the problem remain to be solved by future ^'nvesti- 
 gations. One of the most important questions is the successful com- 
 bination of a positive ferric arsenate with a spreader to give the mixture 
 both spreading and adherent properties. Organic materials which pro- 
 duce a film of the spray on the surface of waxy leaves should have 
 no influence on the electric charge, but the protein type of spreader may 
 have a decided effect. Since most emulsoids, such as gelatin, casein, and 
 proteins, act as protective colloids, their addition to the spray might 
 result in the complete loss of the electric charge. On the other hand, 
 a mutual precipitation may occur (Bancroft, 1920), or if the positive 
 charges are in excess the material may retain its positive character. 
 Only field tests will solve these problems. 
 
46 TECHNICAL BULLETIN 2 
 
 LITERATURE CITED 
 
 Ariial, A. Preparation des bouillies cupriques mouillantes et adherents par I'addi- 
 
 tion de caseine. In Rev. Maladies des Plantes, v. 2, No. 2, p. 9-10, 1914. 
 
 Abst. Rev. App. Ent. v. 2, ser. A, pt. 12, p. 681, 1914. 
 Astruc, H. Bouillies mouillantes. In Prog. Agr. et Viticole, v. 59, No. 24, p. 746- 
 
 753; No. 25, p. 780-782. 1913. 
 , Couvergne, A., and Mahoux, J. Sur 1' adherence des bouillies 
 
 insecticides a 1' arseniate de plomb. In Compt. Rend. Acad. Sci. v. 152, p. 
 
 1860-1862. 191 1. 
 
 Ballard, W. S. and Volck, W. H. Apple Powdery Mildew and Its Control in 
 
 Pajaro Valley. U. S. Dept. of Agr. Bui. 120, 1914. 
 Bancroft, W. D. Mutual Action of Sols. In Jour. Phys. Chem. v. 24, No. i. 
 
 p. 2'i-29. 1920. 
 
 Biltz, W. Ueber die Einwirkung arseniger Saure auf "frisch gefalltes Eisen- 
 
 hydroxyd." In Berichte der Deut. Chem. Gesell. Ja. 2,"^, v. 3, p. 3138-3150. 
 
 1904. 
 Bourcart, E. Insecticides, Fungicides, and Weed-Killers. Scott, Greenwood and 
 
 Son, London, 1913. Translated by D. Grant. 
 Bradley, C. E. Commercial Fertilizers and Insecticides. Ore. Agr. Exp. Sta. 
 
 Bui. 107. 1910. 
 Cazeneuve, — . An Albuminous Bordeaux Mixture. In Vigne France, No. 4, 
 
 P- 51-53- 1898. Cited from Exp. Sta. Rec. v. 10, No. 5, p. 457. 1898. 
 Chapaz, G. Les bouillies mouillantes. Bui. Agr. Algerie et Tunise, v. 19, p. 187- 
 
 191. 1913. 
 Chappaz, G. Les bouillies mouillantes. In Prog. Agr. et Viticole, v. 59, No. 16, 
 
 p. 487-491. 1913. 
 Cooper, W. F. and Nuttall, W. H. The Theory of Wetting and the Determina- 
 tion of the Wetting Power of Dipping and Spraying Fluids Containing a 
 
 Soap Basis. In Jour. Agr. Sci. v. 7, pt. 2, p. 219-239. 1915. 
 Crouzet, E. Bouillie Lacto-cuprique. In La Messager Agricole v. 4, ser. 10, p. 
 
 306-307. 1899. 
 Dalmasso, G. La lotta contro le Tignuole dell 'uva. In Stazioni sperim.entali 
 
 agrarie Italiane, v. 43, No. 7-9, p. 593-645- iQio. 
 Del Guercio, G. I resultati delle prime esperienze tentate con i Polisolfuri colloi- 
 
 dati contro la "Bianca-rossa" delgi Agrumi. In Rivista di Patologia Vegetale, 
 
 V. 7, No. 5, p. 129-135. 1914. 
 Edwardes-Ker, D. R. The Addition of Soft Soap to Lead Arsenate for Spray- 
 ing Purposes. Jour. South East Agr. College (Wye, Kent). No. 22, p. 
 
 359-362. 1913. 
 Fairchild, D. G. Experiments with Fungicides to Prevent Leaf Blight of Nursery 
 
 Stock. Jour. Mycol. v. 7, No. 4, p. 338-353- 1894. 
 Fernald, C. H. A New Insecticide. Arsenate of Lead. Mass. Agr. Exp. Sta. 
 
 (Hatch) Bui. 24, p. z-"]. 1894. 
 Galloway, B. T. Report on Experiments Made in 1891 in the Treatment of 
 
 Plant Diseases. U. S. Dept. of Agr., Div. Veg. Path. Bui. 3, ^6 pp. 1891. 
 . Field Work of the Year in Treating Plant Diseases. In Rept. 
 
 U. S. Sec. of Agr. p. 216-234. 1892. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 47 
 . Some Observations on New and Old Insecticides and Their Com- 
 
 bination with Fungicides. In Insect Life, v. 7, No. 2, p. 126-132. 1894. 
 
 Gastine, G. Sur 1' emploi des Saponines pour la preparation des emulsions insecti- 
 cides et des liqueurs de traitements insecticides et anticryptogamiques. Compt. 
 Rend. Acad. Sci., v. 152, p. 532-534. 1911. 
 
 ■ . Le Progres agricole et viticole, v. 2)3, No. 14, p. 427-429. 191 1. 
 
 Gillette, C. P. Experiments with Arsenites. Iowa Agr. Exp. Sta. Bui. 10, p. 
 401-420. 1890. 
 
 Girard, Aime. Recherches sur 1' adherence aux feuilles des plantes, et notamment 
 auz feuilles de la pomme de terre, des composes cuivriques destines a combattre 
 leurs maladies. In Comp. Rend. Acad. Sci., v. 114, No. 5, p. 234-236. 1892. 
 
 Graham, S. A. Potato Spraying in Minnesota. In 17th Rept. State Ent. of 
 Minn. p. 21-31. 1918. 
 
 Gray, G. P. Wettable Sulfurs. In Monthly Bui. St. Comm. of Hort. Calif., v. 7, 
 No. 4, p. 191-192. 1913. 
 
 Grimaux, E. Sur des sels ferriques colloidaux. Compt. Rend. Acad. Sci., v. 98, 
 No. 25, p. 1540-1542. 1884. 
 
 Guillon, G. AI. and Gouirand, G. Sur I'adherence des bouillies cupriques utilisees 
 pour combattre les maladies cryptogamiques de la Vigne. Compt. Rend. Acad. 
 Sci., V. 127, No. 4, p. 254-256; No. II, p. 423-424. 1918. 
 
 Harkins, W. D. and Brown, F. E. The Determination of Surface Tension 
 (Free Surface Energy), and the Weight of Falling Drops: The Surface 
 Tension of Water and Benzene by the Capillary Height Method. Jour. Amer. 
 Chem. Soc, v. 41, No. 4, p. 499-524. 1918. 
 
 Harkins, W. D., Brown, F. E., and Davies, E. C. H. The Structure of the 
 Surfaces of Liquids, and Solubility as Related to the Work Done by the 
 Attraction of Two Liquid Surfaces as They Approach Each Other. Jour. 
 Amer. Chem. Soc, v. 39, No. 3, p. 354-364. 1917. 
 
 Harkins, W. D., Davies, E. C. H., and Clark, G. L. The Orientation of Molecules 
 in the Surface of Liquids, the Energy Relations at Surfaces, Solubility, Ad- 
 sorption, Emulsification, Molecular Association, and the Effect of Acids and 
 Bases on Interfacial Tension. In Jour. Amer. Chem. Soc, v. 39, No. 4, p. 
 541-596. 11917- 
 
 Hartzell, F. Z. The Influence of Molasses on the Adhesiveness of Arsenate of 
 Lead. In Jour. Econ. Ent., v. 11, No. i, p. 62-66. 1918. 
 
 High, M. M. Cactus Solution as an Adhesive in Arsenical Sprays for Insects. 
 
 U. S. Dept. of Agr. Bui. 160. 1915. 
 Holloway, T. E. The Poison Exponent : A Symbol of the Toxicity of Chemicals 
 in Their Relation to Insects. In Jour. Econ. Ent., v. 5, No. 6, p. 452-456. 
 1912. 
 Holmes, H. N., and Arnold, R. The Peptization of Ferric Arsenate and Phos- 
 phate and the Formation of Their Gels. In Jour. Amer. Chem. Soc, v. 40, 
 No. 7, p. 1014-1019. 1918. 
 
 , and Fall, P. H. The Influence of the Age of Ferric Arsenate on 
 
 Its Peptization. In Jour. Amer. Chem. Soc, v. 41, No. 5, p. 7I3-7I7- IQIQ- 
 -, and Rindfusz, R. E. The Colloidal Arsenates and Phosphates of 
 
 Iron. In Jour. Amer. Chem. Soc, v. 38, No. 10, p. 1970-1982. 1916. 
 Howe, R. G. Results of Spraying Experiments, 1909. HI. Agr. Exp. Sta. Circ. 
 137. 1910. 
 
48 TECHNICAL BULLETIN 2 
 
 Issleib, — . Die Beseitigung der Insekten, welche den Wein- und Obstbau schadi- 
 
 gen, durch Verklebung mit Hilfe von Moosschleim. In Zeitschr. fur 
 
 Pflanzenkrankheiten, v. 24, pt. 2, p. 78-79. 1914. 
 Jacobson, C. A. Alfalfa Saponin. Alfalfa Investigation VII. In Jour. Amer. 
 
 Chem. Soc, v. 41, No. 4, p. 640-648. 1919. 
 Jones, P. R. Machine Gun Work with a New Formula on Red Spider in Tulare 
 
 County. In Monthly Bui. St. Comm. Hort., Calif., v. 7, No-. 7, p. 455-457. 
 
 1913- 
 Kilgore, B. W. On the Cause and Prevention of the Injury to Foliage by Arsen- 
 
 ites Together with a New and Cheap Arsenite and Experiments on Combining 
 
 Arsenites with Some Fungicides. N. C. Agr. Exp. Sta. Bui. 77b. Tech. Bui. 
 
 No. 2. 1891. 
 King, P. E. General Review and Bibliography of Dyeing, ist Rept. on Colloidal 
 
 Chem. and Its Indust. App., British Assoc. Adv. Sci., p. 20-38. 1917. 
 Kirkland, A. H. On the Value of Glucose in Spraying. In Mass. St. Bd. of Agr. 
 
 46th Ann. Rept., p. 478-481. 1898. 
 
 Lafforgue, G. Les bouillies cupriques mouillantes. In La vie Agricole, v. 3, p. 
 
 611-614. 1913. 
 Langmuir, I. The Constitution and Fundamental Properties of Solids and 
 
 Liquids. II Liquids. In Jour. Amer. Chem. Soc, v. 39, No. 9, p. 1848-1906. 
 
 1917. 
 Lavergne, G. Rapport sur le black-rot dans I'Armagnac en 1895. In Bui. Alin. 
 
 Agr., France, v. 15, No. 2, p. 285-291. 1896. 
 Lees, A. H. Winter Cover Washes. In Ann. App. Biol., v. i. No. 3-4, p. 351- 
 
 364- 1915- 
 . Winter Cover Washes. In Ann. App. Biol., v. 2, No. 4. p. 245- 
 
 249. 191 6. 
 . Accessory Wetting Substances with Special Reference to Paraffin 
 
 Emulsions. In Ann. App. Biol., v. 3, No. 4, p. 141-149. 1917. 
 Lefroy, Maxwell H. The Psylla Disease of Indigo in Behar. In Agr. Jour. 
 
 India, v. 8, pt. i, p. 1-26. 1913. 
 
 . Insecticides. In Ann. App. Biol., v. i. No. 3-4, p. 280-298. 1915. 
 
 Lovett, A. L. Spreaders for Arsenate Sprays. In Jour. Econ. Ent., v. 11, No. i, 
 
 p. 66-69. 1918. 
 
 . Insecticide Investigations. Ore. Agr. Exp. Sta. Bui. 169. 1920. 
 
 and Robinson, R. H. Toxic Values and Killing Efficiency of the 
 
 Arsenates. In Jour. Agr. Res., v. 10, No. 4, p. 199-207. 1917. 
 Lowe, V. H. Combating the Cottonwood Leaf Beetle. In N. Y. Agr. Exp. 
 
 Sta.- 15th Ann. Rept. p. 543-544. 1896. 
 Mausier, — . Principles d'apres lesquels doivent etre composes les insecticides. In 
 
 Bui. Soc. des Agr., France, v. 63, p. 86-93. 1908. 
 Marlett, C. L. Comparative Tests with New and Old Arsenicals on Foliage and 
 
 with Larvae. U. S. Dept. of Agr. Div. of Ent. Bui. 6, new series, p. 30-35- 
 
 1897. 
 Melander, A. L. The Control of the Codling Moth. Wash. St. Coll. of Agr. 
 
 Bui. 103, p. 4-55. 191 1. , 
 Millardet and Davis. Resultats de divers procedes de traitement sur le developpe- 
 
 ment du mildiou. In Jour. Agr. Pratique, v. 2, No. 50, p. 761-770. 1886. 
 
SPREADING AND ADHERENCE OF ARSENICAL SPRAYS 
 
 49 
 
 Neiils, J. p. The Use of Flour Paste in Lime Sulphur Solutions in the Con- 
 trol of the Citrus Red Spider. In Monthly Bui. St. Comm. Hort Calif 
 V. 2, No. 6, p. 557. 1913. 
 
 O'Kane, W. C, Hadley, C. H., and Osgood, W. A. Arsenical Residues after 
 Spraying. N. H. Agr. Exp. Sta. Bui. 183. 1917. 
 
 Parker, J. R. The Use of Soap to Retard the Settling of Certain Arsenicals. 
 
 Mont. Agr. Exp. Sta. Bui. 86, p. 35-45. 1911.* 
 Parker, Wm. B. Flour Paste as a Control for Red Spiders and as a Spreader 
 
 for Contact Insecticides. U. S. Dept. of Agr. Bur. of Ent. Circ. 166. 1913. 
 Perraud, Joseph. Recerches sur quelques moyens permettant d'augmenter I'ad- 
 
 herence des bouillies cupriques. In Comp. Rend. Acad. Sci v 127 No 22 
 
 p. 876-879. 1898. 
 
 — . Sur une nouvelle bouillie cuprique, plus specialement destinee a 
 
 combattre le black rot. In Compt Rend Acad. Sci., v. 127, No. 23, p. 978-980. 
 
 Pickering, S. U. The Interaction of Metallic Sulphates and Caustic Alkalies. In 
 
 Jour. Chem. Soc, v. 91.92, p. 1981-1988. 1907. 
 Quincke, G. Ueber den Randwinkel und die Ausbreitung von Flussigkeiten auf 
 
 festen Korpern. In Wiedemann's Annalen der Physik und Chemie, neue 
 
 Folge, Band II, No. 10, p. 145-194. 1877. 
 Rapport au Ministre de L'AgricuIture. Sur le traitement du mildiou dans le 
 
 med6c. In Jour. Agr. Pratique, v. 2, No. 49, p. 659-662. 1885. 
 Ravaz, L., and Bonnet, A. Experiences sur le traitement du mildiou. In Ann. 
 
 de I'Ecole Nationale d'Agri. de Montpellier, n. ser., v. 3, No. 2, p. 157-168 
 
 1903. 
 
 Robak, F. Wild Volatile-Oil Plants and Their Economic Importance. I Black 
 Sage; II Wild Sage; III Swamp Bay. U. S. Dept. of Agr. Bur. Plant 
 Indust. Bui. 235. 1912. 
 
 Sanders, G. E., and Brittain, W. H. The Toxic Value of Some Common Poisons 
 Alone and in Combination with Fungicides, on a Few Species of Biting In- 
 sects. In Proc. Ent. Soc. Nova Scotia, No. 2, p. 55-64. 1916. 
 
 Scott, E. W., and Siegler, E. H. Miscellaneous Insecticide Investigations. U. S. 
 Dept. of Agr. Bui. 278. 191 5. 
 
 Scott, W. W. Standard Methods of Chemical Analysis. Van Nostrand Co, 
 N. Y. 1917. 
 
 Selby, A. D.- Spraying for Apple Scab in 1908. In Ohio Agr. Exp. Sta. Circ. 
 87, p. 4-8. 1908. 
 
 Sirrine, F. A. A Spraying Mixture for Cauliflower and Cabbage Worms. N. Y. 
 
 Agr. Exp. Sta. (Geneva) Bui. 144. 1898. 
 Smith, John B. Insecticides. Rept. of Ent. In 28th Ann. Rept. N. J. Agr. Exp. 
 
 Sta., p. 359-373- 1910. 
 . Insecticides. Rept. of Ent. In 31st Ann. Rept. N. J. Agri. Exp. 
 
 Sta. p. 359-373- 1910. 
 Smith, Loren B. Relationship Between the Wetting Power and Efficiency of 
 
 Nicotine-Sulphate and Fish-oil Soap Sprays. In Jour. Agr. Res., v. 7, No. 9. 
 
 p. 389-399- 1916. 
 Sollmann, T, A Manual of Pharmacology, 1917. 
 
50 TECHNICAL BULLETIN 2 
 
 Stearns, L. A. Experiments on the Control of the Oriental Fruit Moth. In 
 Quart. Bui. Va. St. Crop Pest. Comm, v. 2, No. i. 1920. 
 
 Stewart, V. B. Some Important Leaf Diseases of Nursery Stock. Cornell 
 Agr. Exp. Sta. Bui. 358, p. 167-226. 1915. 
 
 Surface, H. A. Calendar for 1905 for Treating Insect Pests and Plant Dis- 
 eases. Monthly Bui. Div. Z06I. Pa. v. 3, No. i, p. 8-30. 1905. 
 
 Swingle, W. T. An Improved Method of Making Bordeaux Mixture. In Jour. 
 Mycol., V. 7, No. 4, p. 365-371. 1894. 
 
 Tartar, H. V. and Bundy, L. A. Arsenic Soluble in Mixtures of Lead Arsenate 
 
 and Soap. In Jour. Indust. and Engin. Chem., v. S, No. 7, p. 561-562. 1913. 
 and Wilson, H. F. The Toxic Value of the Arsenates of Lead. In 
 
 Jour. Econ. Ent., v. 8, No. 5, p. 481-486. I9i5'. 
 U. S. Patent No. 1 166387. Dec. 28, 1915. 
 Van Slyke, L. L., and Urner, F. A. The Composition of Commercial Soaps in 
 
 Relation to Spraying. N. Y. Agr. Exp. Sta. (Geneva) Bui. 257. 1904. 
 Vermorel, V., and Dantony, E. De I'emploi de I'arseniate ferreux contre les 
 
 insects parasites des plantes. Compt. Rend. Acad. Sci., v. 148, No. 5, p. 302- 
 
 304. 1909. 
 . Arseniate ferreux et bouillies cupriques. In Progres Agricole, v. 
 
 51, p. 423-424- 1909- 
 
 Des principes generaux que doivent presider a I'etablissement des 
 
 formules insecticides. In Compt. Rend. Acad. Sci., v. 151, p. 1144-1146. 
 1910. 
 
 . Sur les bouillies anticryptogamiques mouillantes. In Compt. Rend 
 
 Acad. Sci., v. 152, p. 972-974. 1911. 
 
 Bouillie anticryptogamique au savon de cuivre colloidal. In Compt. 
 
 Rend. Acad. Sci., v. 152, p. 1263-1265. 191 1. 
 
 Tension superficielle et pouvoir mouillant des insecticides et fongi- 
 
 cides. Moyen de rendre mouillantes toutes les bouillies cupriques ou insecti- 
 cides. In Comp. Rend. Acad. Sci., v. 154, No. 20, p. 1300-1302. 1912. 
 
 Preparation des bouillies alcalines mouillantes. In Prog. Agr. et 
 
 Viticole, V. 59, No. 24, p. 745-746. 1913. 
 . Pouvoir mouillant et adherence des bouillies. In Prog. Agr. et 
 
 Viticole, V. 59, No. 25, p. 778-780. 1913. 
 . Preparation rapide des bouillies a la caseine. In Progres Agricole, 
 
 V 63, p. 509. 1915- 
 Volck, W. H. Sulphur Sprays for Red Spider. Cal. Agr. Exp. Sta. Bui. I54. 
 
 1903. 
 . The Apple Powdery Mildew in Pajaro Valley. Office of County 
 
 Ent. for Monterey and Santa Cruz Counties, Cal. Special Bui. i, 1909. 
 Waite, M. B. Experiments on the Apple with Some New and Little-known Fungi- 
 cides. U. S. Dept. of Agr. Bur. PI. Indus. Circ. 58. 1910. 
 Washburn, F. L. Practical Work with the Codling Moth and with a Combined 
 
 Insecticide and Fungicide. Ore. Agr. Exp. Sta. Bui. 10. 1891. 
 Weinmann, J. Bouillies mouillantes. hi Progres Agricole, v. 57, p. 709-712. 
 
 1912. 
 Willows, R. S. and Hatschek, E. Surface Tension and Surface Energy and 
 
 Their Influence on Chemical Phenomena. J. and A. Churchill, London. 
 
 1915. 
 Wilson, H. F. Common Insecticides. Wis. Agr. Exp. Sta. Bui. 303. 1919- 
 
^ 
 
LIBRARY OF CONGRESS 
 C7I V\9\Q\ R«^t^ ^^"^ ^