J 35057 The Virginia Journal of Science ir'\TYRANNUS /SKl // * Sw (may 7- ip X4^Agi-i' Volume 2 (New Series), 1951 EDITORIAL BOARD Boyd Harshbarger, Blacksburg . . . . Editor-in-Chief Horton H. Hobbs, Jr., Charlottesville . Technical Editor Mary E. Humphreys, Staunton . . Assistant Technical Editor SECTION EDITORS W. P. Judkins, Blacksburg . . . - . . . Agricultural Science Irving G. Foster, Lexington . Astronomy, Mathematics, and Physics J. Douglas Reid, Richmond . Bacteriology Ladley Husted, Charlottesville . . . Biology William E. Trout, Jr., Richmond . . . Chemistry Francis G. Lankford, Jr., Charlottesville . . . Education N. F. Murphy, Blacksburg . . . . Engineering B. N. Cooper, Blacksburg . . . Geology William Bickers, Richmond . Medical Science Richard H. Henneman, Charlottesville . Psychology B. W. Jarman, Richmond . Science Teachers Walter A. Hendricks, Bethesda, Maryland . Statistics Published by the Virginia Academy of Science CONTENTS No. 1, January, 1951 (Mailed March 24, 1951) The Physics of Elementary Particles . Frank L. Hereford 1 A New Research Tool and Technique for the Biologist . . . Erling S. Hegre 10 Near Balance Rectangular Lattices . - . Boyd Harshbarger 13 A Punch Card System for the Bibliography of Analytical Chemistry . . . . . . Ralph E. Thiers 28 Acidity-Alkalinity in the Alimentary Canal of Twenty Insect Species . J. M. Grayson 46 Additional Notes on Virginia Plants not Previously Reported from Massanutten Mountain Area . Lena Artz 59 Mesodon /lndrewsae normalis (Pils.) in Virginia . . . . Paul R. Burch & Leslie Hubricht 60 News and Notes . . . . . . 62 No. 2, April, 1951 (Mailed April 16, 1951) Development of the Science of Bacteriology ........ Wilson B. Bell 81 Isolation of a Virus Causing Granulosis in the Red-Banded Leaf Roller . Harrietts Block Wasser & Edward A. Steinhaus 91 Prediction of Subsonic Pressure Distributions by the Sound-Space Theory . Robert Wesley Truitt 94 Gastrophyrne carolinensis carolinensis (Holbrook) in South¬ western Virginia . . James A. Fowler & Richard L. Hoffman 101 The Life History of Gerris canaliculatus Say in Virginia (Hemiptera, Gerridae) . Marvin L. Bobb 102 A Report on the Effect of Ethylene Dibromide Soil Treatment on Root-Knot Control, Nodulation, and Yield of Peanuts . . * . Lawrence I. Miller 109 Protective Behavior and Photic Orientation in Aquatic Adult and Larval Two-Line Salamanders, Eurycea h. bislineata x cirri- gera . John Thornton Wood 113 A New Crayfish of the Genus Orconectes from Southeastern Virginia (Decapoda, Astacidae) . . Horton H. Hobbs, Jr. 122 News and Notes . . . 129 Program for the Twenty-Ninth Annual Meeting of the Virginia Academy of Science . . . . . . 143 in No. 8, July, 1951 (Mailed September 11, 1951) A Significance Test for Differences between Ranked Treatments in an Analysis of Variance . . . D. B. Duncan 171 Bottom Fauna and Temperature Conditions in Relation to Trout Management in St. Mary’s River, Augusta County, Virginia . Eugene W. Surber 190 Pseudoconics . . . D. S. Davis 208 Premenstrual Tension . William Bickers 210 Survey of the Intestinal Helminths of Triturus v. viridescens in the Vicinity of Charlottesville, Virginia . Catherine M. Russell 215 In Memory of John Clayton, Botanist of Colonial Virginia . . Jessie Hopkins & A. B. Massey 220 News and Notes . 220 No. 4, September, 1951 (Mailed October 29, 1951) Proceedings of the Virginia Academy of Science, 1950-51 Minutes of the Twenty-Ninth Annual Meeting, May 10, 11, 12, 1951 Detailed Table of Contents . 242 iv SUBJECT INDEX Acroneuria abnormis . 197 depressa . 197 lycorias . 197 Aeschna . — - 196 Alimentary canal (insect) . 46 Alloperla .... . 197 Analysis of Variance . 171 Anisota senator ia . 51 Annual Meeting, 29th . 245 Notes . 145 Program . - 143 Archips fervidana ... . — 51 Argyrotaenia velutinana . 91-93 Arphia xanthoptera . 52 Atomic particles . 1 Awards Jefferson Medal . 244 J. Shelton Horsley . 68, 244 Bacteriology, History . 81 Baetis cingulatus . 197 Bergoldia clistorhabdion, sp. nov . 92 Brachycentrus . — . 196 Brachycoelium hospitale .... 216-219 Calliphora . 47 Cambarus b. bartonii . 197 Capnia . — § . . 197 Ceratomia catalpe . 51 ChoAmadaplnne calyculata . 59 Chelone obliqua . . 223 Chemistry, analytical Punch card system . 28 Cheumatopsyche . 196 Chironomus . 47, 56 Chloride metabolism . 212 Chortophaga viridifasciata . 53 Chrysomyia . - . 47 Clayton, John . 220 Claytonia virginica . 223 Coleoptera, St. Mary’s R . 196 Committee, Membership . 236 Committee, Reports . Collegiate Members . 276 Fauna . 270 Flora . 269 Foreign Students Exchange .. 266 Finance & Endowment . 258 James River Project . 276 Journal . „ . 271 Junior Academy . 259 Local Arrangements . 63 Long Range Planning . 257 Meeting Place for 1952 . 275 Membership . 267 Research . 68, 257 Resource-Use Education . 268 Science Talent Search . 264 Speakers Bureau . 263 Conics . 203 Contributors, Suggestions for .. inside back cover Crayfish . 122, 197 Datana angusii . 50, 51 Desmognathus . 113 Diplodiscus temper atus . 217 Diptera, St. Mary’s R . . . 196 Dissosteira Carolina . 53 Dolophilodes distinctus . 196 Ephemerella subvaria . 197 Errata . . . . . . 232 Estigmene acraea . 92 Eurycea b. bislineata . 113-120 Eurycea b. cirrigera . 118-120 Financial Statement Academy . 251 Journal . 273 Frog, Virginia . . 101 Gambusia af finis holbrookii .. 115 Gastrophyrne c. carolinensis .. 101 Gerris canaliculatus . 102 insperatus . 102 marginatus . 102 Granulosis . | . 91 Glossina . 47 Goera calcarata . 196 Heliothis armigera . 51 Helminths in Newt . 215 Heterodera . . 110 Hippiscus rugosus . 52 Hydropsyche . 196 Hyhantria cunea . . 51 Hypoderma lineatum . 47 Institute for Scientific Research . 274 Isonychia sadleri . 197 Isoperla bilineata . 197 similis . 197 Junonia coenia . 92 Lanthus . 196 Lattices, near balance rectangular . 13 Leuctra . . . ... . 197 Lucilia . . . . . . 47, 56 Lucilia sericata . 47 Macremphythus varianus . 51 Manuscripts, suggestions for .... inside back cover Megalodiscus rankini . 216-219 Melanoplus bivittatus . 50, 52 differential is - - .... 52 femur-rubrum . 52 Meloidogyne on peanuts . Ill Membership List Senior . 381 Student . 403 Memoriam, In . 284 Davis, Donald Walton . 62 Rudd, Wortley Fuller . 129 v Smith, Henry Louis . 132 Snoddy, Leland Bradley .... 69 Mesodon andrewsae normalis .. 60 Microtechnique . 10 Minutes Academy Conference . 249 Collegiate Luncheon . 281 Council . 63, 248, 286, 288 Dinner Meeting . 283 Junior Academy Banquet .... 290 Junior Academy Meeting .... 290 Sections . 294 Musca . . 47 Neophasganophora capitata .... 197 Neuroptera, St. Mary’s R . . 196 News and Notes . 62, 129, 220 Nigronia serricornis . 197 Ncdulation, Peanuts . 109 Odonata, St. Mary’s R . 196 Officers and Committeemen .... 236 Orconectes juvenilis . 122 limosus . 122 propinquus jejfersoni . . 125, 127 p. propinquus .... 125, 127 p. sanborni . 125, 127 obscurus . 125, 127 rusticus forceps . . 122 virginiensis, sp. nov. . 122-127 Osmunda claytoniana . 223 Par adept ophlebia guttata . 197 mollis . 197 Peanuts . 109 Perla hastata . 197 Peltoperla arcuata . 197 Peridroma margaritosa . 92 Periplaneta americana . 52 pH, Alimentary Canal, Insects . 46 Phlox maculata var. pyramidalis . 61 Pieris brassicae . . . 51 rapae . 51-55 Plagitura parva . 216-219 salamandra . 216-219 Plecoptera, St. Mary’s R . 196 Premenstrual Tension . 210 Pressure, Subsonic distribu¬ tions . - . 94 Presidents, List . 235 Proceedings, 1950-51 . 233 Contents . 242 Protoparce quinquemaculata .. 51 Pseudocleon Carolina . 197 Pseudoconics . 203 Amelia . 207, 208 Boccanegra . 206 Cavaradossi . 205, 206 La Tosca . 204 Psychoda . . 47, 56 Pteronarcys proteus . 197 Punch Card System . 28 Quercus imbricaria . . 59 ixana clamitans . . 117 pipiens . 117 Rectangular lattices . 13 Analysis of variance . 25 Red-Banded Leaf Roller ........ 91 Registration, 29th Ann. Meeting . 287 Rhyacophila fuscula . 196 torva . 196 Root-knot on Peanuts . 109 Sabulodes caberata . 92 Salamander, Protective be¬ havior . 113 Schistocerca americana . 52 Secretary-Treasurer Report .... 249 Sections Agricultural .. 131, 151, 225, 294* Astronomy, Mathematics, and Physics .... 69, 131 153, . 226, 302 Bacteriology . . 69, 154, 306 Biology . 70, 133, 155, 310 Chemistry 72, 133, 158, 227, 321 Education . 74, 135, 161, 332 Engineering .... 74, 136, 162, 335 Geology . . 139, 164, 344 Medical Sciences .... 75, 166, 353 Psychology .... 75, 141, 168, 229, . 364 Science Teachers . 170, 373 Statistics .. 77, 141, 170, 230, 375 Significance Test . 171 Snail . .. . .. . 60 Sound-space Theory . 94 Sparagemon planum . 52 St. Mary’s River . 190 Chemical Data . 192 Physical Data . 192-195 Biological Data . 196 Trout Food . . 199 Stenonema fuscum . 197 varium . 197 pulchellum . 197 heterotarsale . 197 Subsonic Pressure . 94 * Italic numbers indicate abstracts. Syrbula admirahilis . 52 Talinum teretifolium . 59 Test Consistent Significance Levels . 177 Multiple Comparisons . 178 Requirements of Satis¬ factory . 175 Varietal Trials . 187 Trichoptera, St. Mary’s R . 196 Trientalis Americana . 61 Triturus v. viridescens .... 215-219 Trout Management . 190 Variance, Analysis of . 171 Viola conspersa . ; . 59 Virus, Granulosis . 91 Water Metabolism . 212 Z ootermopis angusticollis . 47 AUTHOR’S INDEX Artz, Lena . 59 Bell, Wilson B . 81 Bickers, William . 210 Bobb, Marvin L . 102 Burch, Paul R. . 60 Davis, D. S . 203 Duncan, D. B . 171 Fowler, James A . 101 Grayson, J. M . 46 Harshbarger, Boyd . 13 Hegre, Erling S . i . 10 Hereford, Frank L . 1 Hobbs, Horton H., Jr . 122 Hoffman, Richard L . 101 Hopkins, Jessie . . . 220 Hubricht, Leslie . 60 Massey, A. B . 220 Miller, Lawrence 1 . 109 Russell, Catherine M . 215 Steinhous, Edward A . 91 Surber, Eugene W . 190 Thiers, Ralph E. . . 28 Truitt, Robert Wesley . 94 Wasser, Harriette Block . 91 Wood, John Thornton . 113 vii THE VIRGINIA JOURNAL OF SCIENCE A JOURNAL ISSUED QUARTERLY BY THE VIRGINIA ACADEMY OF SCIENCE Vol. 2, New Series JANUARY, 1951 No. 1 Vol. 2, New Series January, 1951 No. 1 THE VIRGINIA JOURNAL OF SCIENCE Published Four Times a Year: In January, April, July and September, by The Virginia Academy of Science Printed by the Commonwealth Press, Inc., Radford, Va. CONTENTS Pages The Physics of Elementary Particles — Frank L. Hereford _ _ 1-9 A New Research Tool and Technique for the Biologist — Erling S. Hegre.... 10-12 Near Balance Rectangular Lattices — Boyd Harshbarger.. . 13-27 A Punch Card System for the Bibliography of Analytical Chemistry — Ralph E. Thiers . 28-45 Acidity-Alkalinity in the Alimentary Canal of Twenty Insect Species — J. M. Grayson . 46-59 Mesodon Andrewsae Normalis (Pils.) in Virginia Paul R. Burch and Leslie Hubricht 60-61 News and Notes . 62-78 EDITORIAL BOARD Boyd Harshbarger, Editor-in-Chief Horton H. Hobbs, Jr., Technical Editor Mary E. Humphreys, Assistant Technical Editor Clinton W. Baber, Advertising Manager section editors T. J. Nugent Irving G. Foster J. Douglas Reid Ladley Husted William E. Trout, Jr. Francis G. Lankford, Jr. D. H. Pletta Nelson Cooper William Bickers Richard H. Henneman B. W. Jarman Entered as second-class matter January 15, 1950, at the post office at Blacksburg, Virginia , under the Act of March 3, 1879. Subscription — $3.00 per volume . Published at Blacksburg , Va . Mxmlbb "FHBiwmw 01, lOfrl Mailed March 24, 1951 Donald Walton Davis THE VIRGINIA JOURNAL OF SCIENCE Vol. 2, New Series No. 1 THE PHYSICS OF ELEMENTARY PARTICLES Frank L. Hereford Rouss Physical Laboratory • University of Virginia Physical science from its earliest years has busied itself in an attempt to understand the structure of matter. Although the discovery and development of more specialized branches of physics have frequently held a brief spotlight, the structure of matter and the properties of its subunits have consistently been of foremost interest. The past decade has been no exception. Spurred on by the perplexing but fascinating prob¬ lems in the world of elementary particles and by the advent of nuclear power, the physicist has brought the problem to its greatest peak of public and scientific interests. In a recent survey1 of the present occupa¬ tions of those physicists receiving their doctorates between 1936 and 1948, approximately half of those reporting a field of concentration listed nuclear, atomic, or molecular physics. It is this phase of contem¬ porary physics which will concern us here with particular emphasis being placed upon elementary particle physics. The outstanding advances recently made in other and borderline fields are by no means less interesting or less important. Biophysics, solid state physics, microwave spectroscopy, fluid dynamics — all of these fields boast significant achievements in the contemporary picture. Their development and the difficulties which they now face, however, are not markedly different from those in the physics of elementary particles. In his attempt to break matter down into its ultimate subunits, the physi¬ cist has discovered or created a need for no less than 12 such units in the form of charged or uncharged particles of finite or zero mass. (These so-called elecentarv particles are listed in Table I.) The development which has brought about this current state of affairs is an interesting and instructive chapter in the history of physics. Early Discoveries — The first milestones were the discovery of radio¬ activity and the identification of the electron as a particle carrying a negative electric charge. The phenomenon of radioactivity was discovered Editor’s Note: In continuing- the policy established by the Editorial Board to present in the first three numbers of each Volume of the Journal an article of general interest and progrss in the several fields of science, we have invited Professor Frank L. Hereford of the Bouss Physical Laboratory, University of Virginia, to contribute this article from the field of Physics. 1. Bernard C. Murdach & Marsh W. White, Physics Today, vol. 3, no. 9 (1950) pg. 20. APR 2 " 1981 2 The Virginia Journal of Science [January in 1896 by Becquerel; the electron was identified through the work of several men, J. J. Thomson in particular, at about the same time. We now know that the electron constitutes one of the products of radioactive emission, the beta-particle. The other entities sometimes emitted are alpha-particles (Helium nuclei) and gamma-rays (high energy photons). The photons can be thought of as the carriers of the energy associated with light waves, each carrying a very small quantum of energy. Con¬ sidered as particles, they have zero mass and, of course, always travel at the velocity of light, usually denoted by c. The next important contribution came in 1913 with Rutherford’s formulation of the “nuclear model” of the atom, in which a heavy nucleus contains most of the mass and all of the positive charge. This central particle is surrounded by a sufficient number of negatively charged elec¬ trons to exactly compensate for the nuclear charge, Z. Thus Z is an inte¬ gral multiple of the electronic charge. A second important nuclear parameter is the nuclear mass usually given as the total atomic weight, A. Beyond the recognition that A was usually equal approximately to an TABLE I Elementary Particles Name Mass Charge Remarks Photon or gamma-ray 0 O Quantum of radiation energy Electron me = 9 x 10~28gm. — e Also called beta-particle Positron me +e Also called beta-particle Proton 1840 me +e Hydrogen nucleus Neutron 1842 me O Neutrino O O Not yet observed /x — meson 200 me ±e Decays in 2 x 10-8 sec. t r — meson 300 me ±e Decays in 10~8 sec. Neutral meson ? O Existence tentative r — meson 1000 me ? Existence doubtful A — meson 10 me ? Existence doubtful Negative proton 1840 me — e Hypothetical integral multiple of the atomic weight of the hydrogen atom, little was known of the properties of the nucleus at this time. 2951] Physics of Elementary Particles 3 During the following several years it became increasingly apparent that the laws and methods of classical Newtonian mechanics were totally incapable of coping with atomic phenomena, let alone specifically nuclear phenomena. We employ “atomic” to denote the physics of that portion of the atom exterior to the central nucleus, i. e., having to do with the orbi¬ tal electrons and their motions and interactions. Quantum, Mechanics — It was at this stage in 1925 that the bril¬ liant work of Bohr, Heisenberg, Schroedinger, and de Broglie culminated in the advent of quantum mechanics. This development was born of the realization that the very language of classical mechanics was largely meaningless in the domain of atomic phenomena. Classical physics at¬ tempted to specify the “state” of a physical system through the instan¬ taneous values of the positions and momenta of all particles in the sys¬ tem. This procedure we now know to be quite meaningless, because any attempt to devise even an idealized experiment measuring simultaneously the precise values of position and momentum of a particle fails. Quantum mechanics tells us more specifically that the uncertainties in experimental values of position and momentum are related by the Heisenberg Uncer¬ tainty Relation, Ap . Ax = h/4? t (1) where we have written p for momentum, x for position and h for Planck’s constant. This remarkable constant had been introduced previously in physics in Planck’s investigation of radiation phenomena. Its numerical value of 6.6 x 1CT27 erg-sec. is sufficiently small to render the limitations imposed by Eq. (1) negligible in macroscopic phenomena; however, in the physics of elementary particles it is a serious limitation. In any case it appeared that a “physical state” as defined through the classical concept of the simultaneous values of the position and mo¬ mentum of a particle had no basis in reality at all ! The physicist had been obtaining theoretical results expressed in a language through which they could never bo experimentally verified. In alleviating this difficulty quantum mechanics shifted physics to foundations involving strictly observable quantities. The classical equations relating supposedly precise values of physical variables were replaced by a“wave equation” as fol¬ lows: Instead of ^associating a set of values with a physical variable, a mathematical operator was so associated, for example, an operator (which we shall call H) with the energy of the system. The physical state of the system was represented by a complex valued function, 'k. Then the wave equation was written as, I I \J/ = En and interpreted in the following way. The values En which allowed solutions of the equation, \!>, with certain “golden” properties were, according to quantum mechanical laws, the “possible values” of the energy of the system. The laws provided further a method of computing APR 2 - 1951 4 The Virginia Journal of Science [January from S& the probability of observing any given value of the energy. Thus in a series of observations the number of times in which a certain value of a variable was observed could be directly compared to the probability of observing that value according to quantum mechanics. It must be emphasized that this seemingly arbitrary procedure had actually been suggested by a sound line of reason based upon a careful analysis of the processes of measurement. The important point is that instead of predicting the outcome of an experiment in terms of (unob¬ servable) precise values of variables., quantum mechanics furnished a set of possible values and a means of computing the probability of observing any one. The importance of this development and the success which it immediately achieved cannot be overemphasized. It represented the first successful attempt to untangle and analyze atomic phenomena on a quan¬ titative basis. Along with the development of quantum mechanics came an increased appreciation of the significance of Einstein’s Theory of Relativity in the atomic domain. In the hands of Dirac and Pauli the union of these two theories brought about even more brilliant success, leading to the precise calculations of various atomic constants, and even forecasting the discov¬ ery of a new elementary particle, the positive electron. The experimental discovery of this particle, which we now call the positron, came within a few years. It was identified by Anderson and his co-workers in examining the cloud chamber tracks of cosmic rays. This remarkable ultra-high energy radiation which falls upon the earth from outer space was to prove a storehouse of new elementary particles in the future. The Nucleus — During this period there had been relatively little interest in purely nuclear phenomena. Although the disintegration of the nitrogen nucleus had been accomplished by Rutherford in 1919 by alpha- particle bombardment, significant progress seemed to stall until the dis¬ covery by Chadwick in 1932 of the neutron, a particle of approximately the same mass as the hydrogen nucleus (or proton) but carrying no electric charge. In the same year Cockroft and Walton produced nuclear disintegrations with artificially accelerated charged particles (Ruther¬ ford’s projectiles had been radioaetively emitted alpha-particles). The development of the cyclotron and Van de Graaff Generator followed and produced much higher energy beams of charged particles to be used in probing the nucleus. Accurate measurements of atomic weights led to information concerning nuclear structure and to the realization that pro¬ tons and neutrons were the nuclear constituents. The amazing properties of the nucleus gradually became apparent. In this region of space, hardly larger than that occupied by the electron, protons and neutrons of mass some 1850 times that of the electron are tightly packed in condensed phase. The tremendous forces effective there 2952] Physics of Elementary Particles 5 derive from a conversion of a portion of the protons’ and neutrons’ masses into binding energy. The famous relation of Einstein’s E = Me2 (2) stating the equivalence of mass and energy is indispensable in a quanti¬ tative analysis of nuclear phenomena. The exact nature of these nuclear forces remains today among the foremost unsolved problem in physics. During the early 1930’s a number of clues were uncovered bearing upon the problem. The nuclear forces were apparently short range forces, the nuclear particles only showing attraction within about 2 x 10“13 cms. of one another. The total binding energy of a nucleus was obtained through Eq(2) as the energy equivalent to the difference between the observed atomic weight and the sum of the masses of its protons and neu¬ trons in separate unbound states. This binding energy was found to be proportional to the number (N) of protons and neutrons in the nucleus, which fact suggested that the forces had a saturation property. For if a nuclear particle were attracted strongly to every other particle in the nucleus the number of interacting pairs and hence the total binding en¬ ergy should be proportional to N2. Apparently the interaction of a single particle became saturated after interaction with a limited number of its neighbors. No known type of force had these properties. At about this same time the nuclear physicist was faced with another serious difficulty. Nuclear data had indicated that only protons and neu¬ trons existed in the nucleus and that electrons could not by any means be among the constituents. One reason for this followed from the uncer¬ tainty relation. If an electron were confined within a nucleus to a Ax = 2x 1CT13 cm., then the uncertainty in the momentum according to Eq. (1) would prevent its being held within the nucleus by forces of the magnitude which had been indicated by experimental data. Such was not the case for protons or neutrons. As a result of their greater mass, their uncertainty in momentum, AMv, corresponded to an energy much less than that for the electron, and in fact less than the nuclear binding ener¬ gy. Thus these heavier particles could be bound in such a small region of space while electrons could not. The difficulty was that although denied an existence in the nucleus, electrons had been observed to emerge in the form of radioaetively emitted beta-particles. This process is generally referred to as beta-emission. A further difficulty arose in connection with the balance of energy in this process. Whereas the beta-particles were emitted with, energies distributed over a wide range, the loss of total energy of the nucleus in the process was constant from one emission to another. The Conservation of Energy, a prize principle of the physicist, seemed threatened. To re¬ move both beta emission difficulties in a single blow, Fermi proposed a solution which also provided a theory of nuclear forces. Fermi's Field Theory — Fermi proposed that in the process of inter¬ acting with one another the nuclear particles exchanged or shared elec- 6 The Virginia Journal of Science [January tron-neutrino pairs with one another. Thus a new type of particle, the neutrino, was introduced. Its role in Fermi's theory will be evident in the following discussion. The electrons and neutrinos were to exist in the nucleus not as distinct particles but in a sort of virtual state in much the same way as photons exist in an electro-magnetic field. In fact the presence of the electron-neutrino pairs in this sense was referred to as an “electron-neutrino field”. This idea can best be understood through anal¬ ogy with the electromagnetic field. Two charged particles act upon one another through their mutual interaction with their electric field. Two nuclear particles were to act upon one another through their mutual interaction with their electron-neutrino field. Suppose that in the exchange process one of these electron-neutrino pairs were to manifest itself as a distinct pair of particles and escape from the nucleus. Would not this event appear as beta-emission? The emission of the neutrino could serve to carry away the missing energy referred to above. Its presence according to Fermi had been missed by the experimental physicist because of the neutrino’s complete lack of observable properties such as charge, mass, or magnetic moment. In fact, this amazingly elusive particle has escaped direct detection up to the present time. Thus the beta-emission difficulties were removed, though in a somewhat artificial fashion. As regards the results of this theory in explaining nuclear forces through the exchange process, many of the properties of nuclear forces were properly predicted. However, the theory was incapable of recon¬ ciling the observed values of a number of variables. When the observed particle masses, probabilities of beta-emission, and other appropriate constants were fed into the mathematical scheme, the computed strength of the nuclear forces due to the coupling of the electron-neutrino field was too small by a factor of about 1012. Yukawa’s Meson — It was at this stage of development in 1935 chat a Japanese physicist, Yukawa, carried the field theory of nuclear forces a step farther. In particular, he modified the coupling field and forced it to yield agreement among the experimental facts. In so doing he found that a new type of elementary particle was required to play the role of the exchanged particle in the interaction between two nuclear particles. According to Yukawa this particle which we now call the meson, should have a mass about 200 times that of the electron and should be radioactive, disintegrating into an electron and a neutrino. The mathe¬ matical formalism of this theory succeeded in predicting the short range and saturation properties of the nuclear forces. In addition through the disintegration of the new particle into an electron and neutrino, the process of beta emission was accounted for. It was possible to compute the probability of beta emission through computation of the probability of a meson disintegrating in the course of its exchange between nuclear particles. 1951 ] Physics of Elementary Particles 7 For the physicist reader, a few details of this scheme will be of interest. In direct analogy with electromagnetic theory one might suppose the potential of the “nuclear force held” to satisfy the wave equation in empty space: a a d> V '-i-sv/sr = o (3) In association with a single nuclear particle at r — 0 a spherically sym¬ metric solution to this equation (instead of omitting the time dependent factor) is, — constant/ r Realizing that this potential lacked the necessary short range property, Yukawa modified Eq. (3) as follows: A % ~ -frSVStf + k2 = O A spherically symmetric solution to this equation is : e_kr Pi .2 "-+J cd CD a CD PS I 7 13 CD CO -F-5 CD 13 3 2 o E-t \ (intra-block) (k 1) (k2 — 2k — 1) by subtraction 1951 ] Near Balance Rectangular Lattices 19 Another method is to list the variety means and then subtract from each the sum of the constants (Cm) for the replication and block to which the variety belongs. CONSTANTS Set I (W-WO [ (k— 2) W+W'] (kn — Tii) = Cu) (5) Set II (W-WO [ (k— 2) W+W'] (kBsi-To,) = C2i) (6) Set III (W-WO [(k— 2) W+W'] (kB3i — T3i) = C..) (7) Set h (W-WO C (k— 2) W+W'] (kBhi - T„.) = Cm) (8) STANDARD ERRORS OF ADJUSTED VARIETY MEANS To test the significance of two varieties, the standard error of their difference is needed. In the Near Balance Rectangular Lattice Designs there are two simple formulas for the standard error depending upon whether two varieties appear together in any one incomplete block or do not appear together. The standard error of the difference between the means of two var¬ ieties occurring together in an incomplete block is /_2_ V kW [ 1 + k— i f w-w' yi k l (k— 2) W+W' JJ (9) The standard error of the difference between the means of two varieties not occurring together in an incomplete block is / 2 r w-w' 1 V/kWL1+ (k-2)W+W'J (10) The two above formulas can be weighted to give the average stand¬ ard error which is / 2 V kW [ 1 + k i w-w' yi k"— k— 1 V (k— 2) W+W' J] (ID NUMERICAL ANALYSIS To illustrate numerically the method, of analysis for the Near Bal¬ ance Lattice Designs, the analysis of a corn variety experiment is pre¬ sented. Since 20 varieties were to be tested in the experiment, a 5x4 Near Balance Lattice was used. The experiment was set up as in Table I and when placed in the field, the varieties were randomized within each block and the blocks within each replicate. 20 The Virginia Journal of Science [ January After the harvesting of the corn, the yields on a per acre basis were tabulated and compiled for computational purposes as shown in Table IV. The upper figure in each plot refers to the variety number while the lower figure is the variety yield in pounds per acre. The size of each plot was 2x10 hills. The total vield of the varieties for the 5 replications are listed in Table V. The calculations for the analysis of variance followed the formulas given in Table III. In order to compute the sum of squares for blocks corrected for varieties, it is convenient to form Tables VI and VII. Table VI shows the difference between 5 Bm and Tm. The block i totals shown in Table VII and the replicate totals are also used in the compu¬ tation for the sum of squares for blocks. 1951 ] Near Balance Rectangular Lattices 21 TABLE IV Yield Of Corn Varieties By Replications Replication 1 Blocks Block Totals 2 3 4 5 (1) 0 66.1 78.7 73.9 83.2 301.9 6 3 1 9 10 (2) 59.5 0 68.6 65.0 87.8 280.9 i i ii 12 14 15 (3) 86.0 97.5 0 97.4 95.2 376.1 , 15 17 18 20 (4) 77.0 76.8 75.5 0 80.7 310.0 1 1 j 21 22 1 | 23 24 (5) 1 89.5 83.7 | | 77.2 | 83.5 0 333.9 Total (Ri) 1602.8 Replication 2 Blocks Block Totals i 1 i 6 | I 1 11 16 21 (i) ! 0 67.6 | 83.1 72.6 71.7 295.0 1 1 j 2 1 12 17 22 (2) | 72.4 0 77.4 71.9 66.4 288.1 1 f 3 8 18 23 | (3) | 67.2 | 67.0 0 74.6 76.6 285.4 1 1 4 9 14 24 i i /: (4) ! 86.0 69.7 100.0 0 75.6 331.3 1 i 1 i 5 10 15 20 (5) I 96.0 93.0 86.7 82.3 0 358.0 Total (R2) 1557.8 22 The Virginia Journal of Science [January Replication 3 Blocks Block Totals 1 1 12 1 1 1 23 | i 9 1 1 20 1 1 1 (1) o | 1 74.2 1 1 j 64.2 60.4 1 1 71.5 1 1 270.3 1 21 r 1 1 1 1 1 18 4 1 1 i 15 1 1 (2) 1 68.6 1 0 1 1 76.9 i 67.4 1 1 67.6 1 1 285.5 1 16 j 2 1 i 1 1 24 ■ 1 1 10 1 1 (3) 1 68.1 j | 63.6 1 1 1 0 1 61.1 1 1 75.7 1 268.5 [ 11 ! | 22 1 1 8 1 ! i 5 1 1 I (4) | | 75.8 | ! 68.8 1 1 66.3 1 0 I 1 83.0 1 1 293 9 i 1 6 J 17 1 [ 3 j 14 1 i i 1 1 1 (5) ! 58.8 | 73.7 71.2 | 86.4 I 1 0 1 290.1 Total (R3) 1403.3 Replication 4 Blocks Block Totals 1 17 8 1 ! 24 1 1 1 15 1 1 1 (1) 0 56.4 53.6 50.4 1 1 60.3 1 1 220.7 16 23 i i i 14 1 1 5 1 1 (2) 57.3 0 57.3 i i 63 9 1 62.9 I 241.4 i 6 j 22 i 1 4 1 1 1 20 1 1 l (3) | 47.0 58.1 0 1 55.6 1 58.7 1 1 219.4 ! ! 21 12 3 ! 1 1 1 10 1 > 1 l (4) ! | 66.3 66 2 61.4 i ! 0 ! 62.9 1 1 256.8 1 1 1 11 1 | 2 18 1 i i 9 1 I i I 1 1 1 (5) 1 68.7 I 51.1 5 I 56 8 i 1 0 1 1 240 1 Total (R.) 1178.4 1951 ] Near Balance Rectangular Lattices 23 Replication 5 Blocks Block Totals 22 18 14 1 10 (1) 0 76.2 72.2 i 87.8 | 75.0 311.2 1 11 3 i 24 j | 20 (2) | I 79.5 j 1 o 75.1 I 68.6 1 84.6 307.8 1 ' 21 i 17 9 1 5 (3) 1 1 80.8 | 78.7 0 62.7 88.9 311.1 I 1 6 1 2 1 1 23 15 j (4) ! 66.8 1 66.4 1 75.9 0 86.0 295.1 1 1 16 | 12 i 8 4 j j (5) 1 68.4 | 77.6 I 70 0 ! 66.3 ! 0 282.3 Total (R5) 1507.5 TABLE V Variety Total Yields Variety Total Variety Total 2 319.6 14 435.5 3 353.6 15 395.8 4 349.2 16 343.4 5 414.0 17 357.5 6 299.7 18 362.7 8 325.5 20 377.8 9 314.6 21 376.9 10 394.4 22 353.2 11 393.1 23 351.2 12 392.9 24 339.2 24 The Virginia Journal of Science [January TABLE VI kBhi — Thi i Tu 5Bn-Tu T2, 5B2i — T2j T3I 5B,‘ji — T3I 1 1436.4 73.1 1413.1 61.9 1436.5 -85.0 2 1334.2 70.3 1423.2 17.3 1484.6 -82.1 3 1617.3 263.2 1393.0 34.0 1396.6 -54.1 4 1441.4 108.6 1438.5 218.0 1485.8 -16.3 5 1420.5 1 ! 249.0 1582.0 208.0 1446.3 + 4.2 i T*i 5Bu — T41 T5i 5B5i — T5i 1 1 1418.0 | | -314.5 1545.8 10.2 2 1544.1 -337.1 1463.7 75.3 3 1379.9 -282.9 | 1463.0 j 92.5 4 1517.8 1 -233.8 | 1366.3 109.2 5 1390.0 j -189.5 1411.0 .5 TABLE VII Block Totals For All Replicates (Bn -f- B21 + Bsi + Bn -J- B5i) i I Total 1 2 3 4 1399.1 1398.7 1460.5 1487.1 5 1504.4 1951 ] Near Balance Rectangular Lattices 25 The analysis of variance table for the experiment is shown in Table VIII. In adjusting the variety means using the inter- and intra-block var¬ iances, the weights and correction terms are calculated from formulas (3) and (4) and formulas ( 5 ), (6), (7) and (8). respectively. W was found to be .062734 and W' to be .013631. The correction terms are listed in Table IX. The unadjusted variety means, the appropriate cor¬ rections applied to each, and the resulting adjusted variety means are shown in Table X. The standard error of the difference between the means of the two varieties (1) occurring in an incomplete block is 2.760 (2) not occurring together in an incomplete block is 2.816. The average standard error is 2.772. TABLE VIII Analysis of Variance of a Near Balance Rectangular Lattice Experiment Source of Variation Degrees of Freedom Sum of Squares Mean Squares Replicates 4 5711.12 1427.7800 Blocks 20 1237.53 61.8765 (P) Varieties (unadjusted) 19 4652.08 244.8463 Error (intra-block) 56 892.65 15.9402 (Q) Total 99 12,493.38 TABLE IX Correction Terms i Cu On Cai C4, Cat 1 .724 .613 -.842 -3.117 .101 2 .697 .171 -.814 -3.341 .746 3 2.608 .337 -.536 -2.804 .917 4 1.076 2.161 -.162 -2.317 1.082 5 2.468 2.061 +.042 -1.878 .005 5 26 The Virginia Journal of Science [January I TABLE X Unadjusted And Adjusted Variety Means Variety Mean Adjustment = Adjusted Mean 2 63.92 — ( Oll-f“C22-]- C33-}" C45-}- Co4 ) = 64.36 3 70.72 — (Cn+C«H-C»H-C44+C») = 71.19 4 69.84 — ( Cll-j-C21-f-C32-j-U43-f"Us5) — 70.57 5 82.80 — ( Cu“f- C25-)- C34-I- 0 12 -f~ C53 ) = 82.60 6 59.94 — ( Cl2“|-021 -j-C35-}~Ol3-l-U34 ) = 60.31 8 65.10 — ( Ol2-J~ C23-I- Oai-]- C41-j~ C 35 ) = 67.34 9 62.92 — (Cl2~|-C24-]-C31-J-C45-}-C33) = 61.86 10 78.88 — ( Cl2+ C25 + C33-I- C44-}- Csi ) = 78.87 11 78.62 — ( Cl3-|- C2I -4- 034-f" C45-}- Co2 ) = 76.69 12 78.58 — (CX3+C22+C3X+C44+C55) = 78.96 14 87.10 — ( Ol3-(- 024-(- O35-}- 042-f“ Col ) = 85.53 15 79.16 — ( Cl3~f~U25-|-U32-]-C41 — j— C54 ) = 77.34 16 68.68 — ' (0l4-J-C21-f'C33-f"042-(_U53) — 70.86 17 71.50 - ( Cl4“|~ ^—^22 | ^—^33 { C4I C53) = 72.41 18 72.54 — ( Cl4-f~ C23-I- C32-j- Cxo-j- Csi ) = 73.72 20 75.56 — ( 0l4-j-L25_j“031_}-L43_i-U52) = 75.32 21 75.38 — ( Oio-j- 02i -j- O32-I- C 11— j— 0o3 ) = 74.51 22 70.64 — (Cl5-J-U22“l-C34 4-C43-f-Cr>1 ) rr 70.87 23 70.24 — (0l5“l-023-|-031-l-042-j-U34 ) = 70.54 24 67.84 — ( Ol5 “(- C24 “j- C33-}- C 41 “J- C.-.2 ) — 66.12 1951 ] Near Balance Rectangular Lattices 27 REFERENCES G. M. Cox, R. C. Eckhart, and W. G. Cochran, 1940. — The Analysis of Lattice and Triple Lattice Experiments in Corn Varietal Tests. Iowa Agri. Exp . Sta., Bill., 281. Boyd Harshbarger. 1947. — Rectangular Lattices. Virginia Agricultural Experiment Station. Memoir 1. Boyd Harshbarger. 1949. — Triple Rectangular Lattices. Biometrics, American Statistical Association, 5(1). F. Yates. 1936. — A New Method of Arranging Variety Trials Involv¬ ing a Large Number of Varieties. Journal Agri. Sci., 26: 424.455. F. Yates. 1939. — The Recovery of Inter-Block Information in Three Dimensional Lattice. Annals of Eugenics, 9 : 136-156. F. Yates. 1940. — The Recovery of Inter-Block Information in Bal¬ anced Incomplete Block Designs. Annals of Eugenics, 10: 317-325. 28 The Virginia Journal of Science [January A PUNCH CARD SYSTEM FOR THE BIBLIOGRAPHY OF ANALYTICAL CHEMISTRY Ralph E. Thiers John Lee Pratt Trace Analysis Laboratory, School of Chemistry, University of Virginia INTRODUCTION At one time or another every scientist has been faced with the great task which is involved in concentrating from the literature the informa¬ tion pertinent to his particular problem. This is a long job, and the amount of information which must be filed in one research laboratory often has such bulk that even to obtain information on one topic is time consuming. The increasing seriousness of this problem as the store of scientific literature grows has been viewed with alarm in many articles and editorials in recent years. The source of the difficulty has been attri¬ buted to the fact that the art of indexing and classifying information has not kept pace with the increase in the amount of information to be indexed, and the old methods are becoming less able to cope with the growing mountains of material. Dr. V. Bush, in the Atlantic Monthly of July, 19451 discussed this subject and suggested that someone perfect a mechanical memory called the “Memex” machine in which vast quan¬ tities of material could be stored and brought out at will with high efficiency, and under many classifications. Some work has been done on such a device, but until it is perfected other methods must suffice. PUNCHED CARDS General The problem has been attacked, and in some phases partially solved by the use of punched cards. These are filing cards with which the opera¬ tions of “sorting” and “selecting” may be done mechanically by virtue of information placed on them as code in the form of holes or slots. The sorting or selecting mechanism operates by responding to this code and accepting or rejecting any card during a given operation according to the position of the holes or slots. The more information a card can carry as code the more valuable it becomes as a unit in a filing system. This is because sorting or selecting can be done in terms of one class of infor¬ mation independent of the others and therefore each type of information 1951 ] Punch Card System for Bibliographies 29 coded can be consulted as though the file were set up for that type. This completely eliminates the necessity for cross referencing. Probably the best known system for making and handling punch cards is that of the International Business Machines Corporation, and it is widely used. However the equipment is too involved and expensive for use by a small laboratory. Another system of punch cards exists which fits very well the needs of a small organization or even an indivi¬ dual. This system involves the use of edge-punched and manually slotted cards, and because of its simplicity, nominal cost, and wide applicability it deserves more extensive use. Such cards can be obtained under the trade name of “Rocket” from The Charles R. Hadley Co., 330 North Los Angeles Street, Los Angeles, California, and under the trade name • • • V • • \MA Tryv • • • • FIG. 1 7 4 2 1 T 7 4 2 1 "9" 7 4 2 1 "2“ 7 4 2 1 "0" FIG. 2 FIG. 7 30 The Virginia Journal of Science [January of “Keysort” from the McBee Company, Athens, Ohio, and have been discussed in several articles2,3,4,R,G. This paper is an attempt to summarize the techniques in use, to generalize on the possibilities for new techniques, and to describe one card form which should be widely adaptable. Methods The basic principle of the edge notched punch card is that when a number of cards have holes punched in a certain place near the edge, and when cards corresponding to a given requirement have a slot cut so that this hole is opened to the edge (as shown in fig. 1) a needle may be passed through one of the holes in a stack of cards and the stack lifted, dropping out only those cards which have been slotted2. Direct Coding — A row of holes may be punched on each card and given different independent meanings, and can be slotted or not slotted in each position in accordance with the criterion for that hole. This has been termed direct coding and is the simplest method in use. If, for example, hole number 3 is slotted when the card is concerned with the element uranium, then by needling hole number 3 all of the cards on uranium can be obtained from the pack. Field Coding , 7Jf21 Type — Ingenuity soon led to improvements in the use of holes to gain versatility and to allow coding of numbers and letters. The idea of using several holes side by side combined in a “field” for coding was the basic step away from direct coding3. Numbers can then be coded by a system utilizing a field four holes wide with the holes labelled from left to right 7, 4, 2, and 1, respectively. By slotting one or more of these holes so that the sum of the numbers slotted represents the desired figure, digits from 1 to 9 can be coded (no slots representing zero), and all numbers up to 14. (See Fig. 2). By using several adjacent fields for the units, tens, hundreds, etc., large numbers can thus be coded. There is, however, a loss in going from the direct coding to the above described field system. In direct coded cards all of the cards corresponding to a certain classification can be obtained by one pass of a needle; or if the cards are desired in a certain order of classifications they may be placed in that order by needling the holes in the appropriate order, which is of course a long business. These two operations are “selecting” and “sorting”, respectively. The simple 7421 field does not allow the selection of numbers, but gains a great advantage in sorting efficiency. It can be seen from the above example (Fig. 2) that if needles were inserted to drop out 1920, then along with 1920 would also come °11 four digit numbers with 1, 3, 5, or 8 as the first digit when followed by 9 as the second and 2, 3, 6, or 9 as the third. However this code does provide the ability to sort the cards into numerical order, which is done 1951 ] Punch Card System for Bibliographies 31 by the following process. A needle is inserted into the hole at the extreme right and any cards which fall are placed at the back of the stack; the hole immediately to the left is now needled and the fallen cards placed at the back; this is continued until all the holes have been needled, when the pack will be found to be in numerical order, with the lowest numbers at the front. Selectable 7.I/.21 Code. — In order to gain the ability to select, as well as sort, two more holes must be added to each field. These holes are labelled 0, and SF, the latter standing for “single figure”. (Fig. 3). Then when 7, 4, 2, 1, or 0 is to be coded the SF hole is also slotted, and when 7, 4, 2, 1, or 0 is to be selected from a certain field two needles are used, one in the figure hole and one in the SF hole. The needle through the SF hole then prevents double figure cards from dropping. Alphabetic Fields. — Coding of letters of the alphabet may be done by two methods, one of which adapts numbers, and the other, letters, to the job of being mnemonic devices3’4. In the numerical system the letters A through M are numbered 1 to 13, and slotted as such in a numbers field. To this field one extra hole is added, labelled N-Z, and the letters N through Z are also numbered from 1 to 13 and so slotted, but with the N-Z hole slotted also. This of course provides sortability but not selectability. (Fig. 4). The mnemonic letter system also uses a field five holes long, the holes being labelled O, I, E, C, and B. Each letter is then assigned a slotting code as follows: A — no hole slotted B — B only slotted C — C only slotted D — C and B slotted E — E only slotted F — E and B slotted G — E and C slotted H — E, C, B slotted, and so on. This also allows sorting of the cards but not selection of cards coded with a certain letter. It may be seen that the two systems are closely related. A new and ingenious system of alphabetic and numerical coding was described by Cox, Casey and Bailey5 which departed from previous practice in using a triangular code. By means of only two slots in a field five holes wide any digit from 0 to 9 could be coded and used either for selecting, or for sorting by the usual method. In each field the two holes are slotted whose diagonal columns intersect at the desired number. To select a given number two needles are inserted in the holes that would be slotted to code that number, and the pack is lifted, whereupon the selected cards drop out. (Fig. 5). This system is superior to the 32 The Virginia Journal of Science [January earlier one in that more than two needles per field are never needed, and the fields are only five holes wide. It was further improved by the same workers by introducing the idea of using a double row of holes in con¬ junction with a triangular code. Two symbols can then be placed in each square of the fields thus greatly increasing the capacity. The desired square is then designated by the position of the slots as with the tri¬ angular code above, and if the upper symbol of the two occupying that square is meant, the left hand slot is to the first row of holes only, while if the lower symbol is desired, the left hand slot is deep. The right hand slot is, by convention, deep when the left slot is shallow, and shallow when the left slot is deep. (See Fig. 6). By this system a numerical field is only four holes wide per digit, and yet it allows sorting, and selecting with one pass of the needles, using only two needles per field for the selection. Sorting is accomplished in a fashion similar to the one described above for codes using a single row of holes. A needle is passed through the upper hole at the extreme right and lifted; the fallen cards are placed at the back of the pack and the needle passed through the lower hole at the right, and the process repeated, needling the upper then the lower hole from right to left until the left hand side is reached, when the cards will be found to be in order. Selecting is done by needling the holes which would be slotted to code the card whose selection is desired. With this system it is possible to make an alphabetic code without resorting to numbers as mnemonics, for with a field six holes wide, fifteen squares, or space for 30 symbols, are available, or if slight abbreviations such as grouping xyz together are not objectionable, a five hole field with 20 spaces may be used. Because of the variation in frequency of occurrence of the letters when they appear as the first, second, or subsequent letters of a word, Cox, Casey and Bailey5, after studying letter frequencies for names, proposed an excellent breakdown of the alphabet, where the code varies for each of the first three letters of the name. With a code six holes wide, thirty positions, or four more than were necessary to code the alphabet were available. These they occupied by dividing the letter M into Mi,: which are M’s before Me, Me itself, and M« for M’s after Me, and by dividing the letter S into Si, for S’s before Sch, Sch itself and Sa for S’s after Sch. They found that in about 80% of names the second letter was one of A, E, I, O, C, FI, L or R and that one of the group C, H, L, R occurred with about the same frequency as any of A, E, I, or O. Therefore they proposed a full code for the first letter, but an abbreviated three-hole-wide code for the second letter. In the case of the third letter a six-hole-wide code was pro¬ posed, but with different alphabetic distribution. Because of the frequent occurrence of E, L, N, and R as third letters, each of these was divided into two categories, Ei, Ea, Li, La, etc., depending on whether the fourth letter of the word was from the first or second half of the alphabet, 1951] Punch Card System for Bibliographies 33 respectively, thus obtaining somewhat finer sorting. (Fig. 7). By this coding system cards may be sorted and selected by the same method described for its numerical analogue. This double-hole, triangular field coding system represents probably the best and most versatile system for numerical and alphabetical coding at present known for edge punched and slotted cards, and allows a large amount of infor¬ mation to be placed on a relatively small card. Other Methods. — Other methods of coding exist, and can be devised. The same authors have proposed an improved six position selector code, one hole deep. They have also suggested that alphabetic sorting may be done by assigning numbers to equal alphabetic intervals. Use of a double row of holes introduces the possibility not only of slotting shallow or deep, but also of cutting out the space between the two holes. However, cards slotted in this fashion do not drop clear of the pack when needled, and so lose to a large extent their advantage of convenience. (Fig. 8). A system basically different from those above has been described by Mooers7 and Wise8, the mathematical theory of which has been dis¬ cussed by the latter9. This difference is that more than one entry can be coded in a single field. Mooers’ method is to use one large field" and to assign to any word to be coded a given number of randomly chosen slotting positions. He states that a four-notch code in a field of 40 positions will allow up to seven coded ideas from a vocabulary of 90,000 coded ideas to be coded on each card. In this system, and that of Wise, extra or unwanted cards drop with those sought, but their number is small enough that they can be rejected by inspection. Wise uses a num¬ ber of fields, with a number of slots in each, and each idea is assigned a code which gives it one slot in each field. He points out that in order to have maximum efficiency the number of fields should be so related to the number of ideas per card and the total number of available holes that on the average 37% of the holes should be slotted. These systems put few limitations on the material which can be coded. As well as words or numbers, chemical formulae and ciphers for organic compounds from the Dyson or the National Research Council system can be used. A full discussion of these codes will be found in a book edited by Robert S. Casey and J. W. Perry and soon to be published by the Reinhold Publish¬ ing Corporation. THEORETICAL POSSIBILITIES FOR NEW METHODS It is interesting to examine the methods of coding, sorting and select¬ ing these cards which are now in use, and to try to devise new and better codes. One might casually guess that the efficiency and versatility of the cards might be improved indefinitely as better and better systems of coding were found. On second thought, however, one realizes that all 34 The Virginia Journal of Science [January of the above described systems are merely mnemonic devices and that a card with a given number of holes punched in it has a maximum to the number of data which can be recorded on it. This maximum should be calculable, and attempting to do so led to the following results. Let: W stand for the width of the field in question expressed as a number of holes; D stand for the depth of the field in question expressed as a number of holes; N stand for the number of holes opened to the edge to code the data in question; (e.g., for a shallow slot N — 1, for a slot 2 holes deep N — 2, and if one of each of these is in the same field, N = 3) ; K stand for the number of needles used for one pass, in selecting; C stand for the available number of different choices of data to be coded when sortability but not selectability is required; S stand for the available number of different choices of data to be coded when both sortability and selectability are required. Direct Codes. — Each hole of a direct code may be considered as a field, where W = 1, D = 1 and N — 1, and therefore S = 1 and C — 2. If all the holes are independent of each other then the number of data which can be coded is equal to the number of holes on the card. It is possible to make the holes partially dependent, for example by deciding that “if hole number one is slotted indicating that the card refers to a book, then hole number three will be punched if it is over 500 pages long, whereas if hole number two is slotted indicating reference to a journal article then hole number three will be punched if it is over 5 pages long.” This type of system can greatly increase the capacity of a direct type code, but from a strict standpoint it is not a true direct code, since the holes are not independent. It really amounts to using part of the possibilities in a field, since when this dependence exists N becomes 2, or else two passes are required for selection. Field Codes — Sortable. — If selectability is not a requirement and sortability is desired then each different way that a field can be slotted represents one choice, and the following expression holds true: C = (D + l)w This is so because one position one hole deep gives two choices, slotted or unslotted; a position two holes deep gives three choices, and so on. This is expressed tabularly within reasonable limits of D, W, and C in Table I. 1951] Punch Card System for Bibliographies 35 TABLE I. Values of C Field Codes — Selectable. — The number of different ways of slotting cards in a certain field so that they are selectable (and therefore also sortable) depends of course on the width and depth of the field, and also upon the value of N, the number of holes opened to the edge, and the number of needles used, K. Some thought will show that for a field of certain size (D and W constant) S can be calculated for various values of N ; and for a certain N value, the maximum value of S equals the sum of S’s obtainable with each value of K possible at that N value. An example will help to explain this. Assume we have a field where D = W — 3, and we wish to use N = 3. Then when K = 1, one of the three rows is slotted three deep and there are three selectable possibilities. When K = 2 one row is slotted two deep and another only one deep providing six selectable choices. Since two needles are involved none of the K = 1 cards will drop. When K = 3 all three holes are slotted one deep and one choice exists, and since three needles are used this does not conflict with any K = 1 or K — 2 choices. (See Fig. 9). It is possible then to calculate tables of choices (S values) for different values of K and N when D and W are constant. Table II gives these values for D = 2, when W = 4, 5, and 6. TABLE II. r w = 4 i r A* I: 1 w = 6 _ 1 1 2 A A E =1 7 A I 12 [3 A A 6 7 8 U ip 1 2 3 A 5 6 7 8 1= a 11 12 Kl K i 1 A 4 f= 5 5 \ 6 6 j 1 2 A 1 2 ¥ ' Ip ID 15 ¥ 15 2 5 A u _ 12 ~4 iP 30 30 10] 30 00 60 30 5 4 A A A A 5 2C 3 20 ¥ isl 60 90 60 4 5 1 a 10 10 T r A 30 60 i ¥ 6 5 6 1 1 6 3D L5 ¥ T~ 1 .6 (Total I T) H w m 1 T T 5" 15 30 45 a 1 W S r T 6 ¥ 6C E W 151 126 H 60 21 6 It can be seen from this table that the maximum values of S occur DW when N is as close as possible to - , i.e., when half the available A holes are slotted. Calculations for other D and W values within the practical range confirm this. With this in mind it is possible to cal- 36 The Virginia Journal of Science [January • • • • • • • • • • • • 9 • • • vJ • • K = I UAT • • • • • • K = 3 u ® « © U © U • • © • © u u © © U ® • • 51 41 31 21 1 1 1 51 41 31 21 1 1 1 52 42 32 22 12 2 52 42 32 22 12 2 53 43 33 23 13 3 53 43 33 23 13 3 54 44 34 24 14 4 54 44 34 24 14 4 55 45 35 25 15 5 55 45 35 25 15 5 56 46 36 26 16 6 56 46 36 26 16 6 57 47 37 27 17 7 57 47 37 27 17 7 58 48 38 28 18 8 58 48 38 28 18 8 59 49 39 29 19 9 59 49 39 29 19 -9 60 50 40 30 20 10 60 50 40 30 20 10 culate and construct a table of maximum S values for different values of D and W. Table III is such a table. It is, in other words, a tabula¬ tion of the maximum number of data which can be placed in fields of different types, no matter what coding system is used, and still meet the requirement of selectability. 1951 ] Punch Card System for Bibliographies 37 TABLE III. Maximum Values of S It is interesting to note that in the alphabetic code of Cox, Casey and Bailey, in which W = 6 and D = 2, only 30 possibilities are utilized, whereas Table III gives a possible maximum of 151 ! By referring to Table II one can see that they are using all of the possibilities where W — 6, D = 2, N = 3, and K = 2. Even with ' N = 3 only half of the possibilities are being utilized, for with K = 3 thirty more are available. It is when N = 6 that S = 151. While it is not difficult to calculate what the cards are able to do it is a task requiring great ingenuity to devise a mnemonic code to take advantage of their possibilities. It is of course not necessary to use a mnemonic; all of the slotting possibilities could be listed in a book with a meaning beside each, and this book referred to during coding or select¬ ing. However, unless one is extremely crowded for coding space on the card the work and time saved by a mnemonic far more than makes up for the loss in efficiency of space utilization. Examples of the type of mnemonic code that might be used to get more from a code where W = 6 and D = 2 have been worked out by the author for N = 4, and N = 5. The former of these is suggested as a simple, useful code, but the latter is given purely as an example. Mnemonic Code for W — 6, D — 2, N = 4 (Fig. 10). — In order to code one of the symbols of the rectangular code of Fig. 10 the following conventions are followed : a) . The column in which the symbol appears is slotted deep. b) . The number of spaces down the column is indicated by imagin¬ ing that the five remaining unslotted spaces are labelled 7421 SF, and by slotting two shallow holes to give numbers from 1 to 11 not including 10. Thus by reading 10 for 11, a code 10 deep and 6 wide may be obtained which gives sixty selectable and sortable choices. While it is impossible to print 7421 SF on the card for this system 38 The Virginia Journal of Science [January (since the location of these symbols changes with that of the deep punch) even a person who is unfamiliar with punched cards quickly grasps the method and after a little practice finds the system very simple. It may be used without actually printing the code on the card if it is considered that the position of the deep punch gives the tens digits on a field marked 5, 4, 3, 2, 1, 0 and the two shallow punches give the units digits on a 7421 SF field comprised of the remaining holes. (See Fig. 11). This is so easy to remember that printing the code is really superfluous. This code is simple and direct and can be utilized with any cards having a double row of holes. If still further possibilities are desired in the same field fifteen more can be obtained by slotting four of the six positions one deep. This might be coded by allowing the unslotted positions to indicate the symbols on a triangular field. (Fig. 12). Since with this code the unslotted positions indicate the symbols, but selecting is done by needling the slotted holes, the numbers or symbols would appear in order opposite to the usual one, as in Fig. 12. Still another fifteen possibilities may be utilized by slotting two positions two deep. The code here might be analogous to that described immediately above, except that the slotted holes would indicate the symbols and therefore the numbers would not be reversed. All of these three methods may be used simultaneously in the field described, and they give the maximum number of selectable choices for W = 6, D = 2, and N — 4; however, to gain simplicity in sorting only the K = 3 or at most the K = 3 plus K = 4 systems might be used. This is because sorting can best be accomplished by needling the top row of holes from right to left in order and placing the fallen cards at the back after each pass, then repeating this for the deeper row of holes. This puts the K := 4 cards ahead of the K= 3 ones, and puts them in order using the codes shown in Figs. 11 and 13, but it mixes the K = 2 cards with the K = 3 cards, thereby causing complication if they (the K = 2 cards) are used. Mnemonic Code for W — 6, I) = 2} N — 6. — While the above code is simple enough to be useful, the example devised for N = 6 in the same field is one where the code is getting beyond the bounds of practicality. The greatest number of choices per K value can be obtained when K — 4, and therefore when two deep and two shallow punches are made. Symbols might then be coded by use of the following conventions: a) . With a triangular frame of 15 spaces (as in Fig. 13) the square of the frame in which the symbol lies is indicated by the position of the two sets of holes slotted two deep. b) . The remaining four sets of holes (two of which are to be slotted one deep) are considered to be a 4, 2, 1, SF, system by which it is possible to indicate any one of six symbols in each of the squares. 1951 ] Punch Card System for Bibliographies 39 This provides the 90 choices predicted for K — 4, and still allows 60 other possibilities when K — 3, 5, and 6. General. — It may be seen from the above discussion that while Table III gives the maximum number of selectable possibilities, a prac¬ tical limit is imposed by the mnemonic code used. Thus the closeness of approach to the maxima of Table III depends on the capabilities of the code, and on the maximum number of needles and slots one is willing to use.( In judging the efficiency of a code it is useful to employ the concept of the number of combinations available per needle used in sorting. This was introduced by Wise and given the notation Cn, but in the notation of g this article it would be — since the C" values for the entries in Table K II can be obtained by merely dividing the numbers in the body of the table by their corresponding K value. It can be seen that for the code of Fig. 10, Cn = 20. This means that from a mechanical point of view this code is better than those where Cn is lower (e.g.. Fig. 7, 30 Cn = — — = 15) and poorer than those where Cn is higher. The question then arises, which of the several above described systems is best to use? The writer’s experience indicates that for alpha¬ betic and numerical fields the triangular codes of Figs. 6 and 7 are unsurpassed, except in cases where a number of mutually exclusive data are to be coded, (such as the journals on the card described below,) or where a finely dividable alphabet is desired. Then the rectangular code of Fig. 10 is probably best. This is of course a matter of opinion, and some users even prefer single depth codes. It is therefore desirable, when deciding on the code one is going to use, to examine all of the above described methods as well as those of Wise, Mooers, etc., and weighing simplicity against efficiency, pick the best ones for the case in hand. Other basic changes in the method of slotting cards may be developed to gain versatility beyond that predicted above, such as the use of slanted slots, where the degree of slant determines the hole’s meaning. Such changes however are difficult to foresee. CARDS FOR BIBLIOGRAPHY OF ANALYTICAL CHEMISTRY Types of Data. — In analytical chemistry the written portion of the cards should generally contain the following information: Author’s name(s) —Last name and all initials Journal of Publication — Full name or authorized abbreviation Date — Volume, page, month, year, and any other necessary sub¬ division Title of article 40 The Virginia Journal of Science [January Element or material for which analysis is being made Material which is being analysed — Form, source Method of analysis — Reagents, apparatus, technique Results of analysis Range over which method is applicable — Sensitivity and upper limit Precision and accuracy of method A fuller discussion of the classification of this material has been given by Patterson and Mellon10 and anyone devising a code is well advised to refer to their work as well as the index entries of Chemical Abstracts and other reference works. The following table indicates the desirability of coding each of these classes of information, (or the usefulness of the coded information) and the type of code required. It is obvious that field type of code can be used only where data are mutually exclusive, that is, it could not be used for elements sought, since more than one element might be sought, but it could be used for journals, since a reference is to one journal per card. With these factors in mind a card has been designed for use in the literature searching, abstracting, and indexing that is carried on in the Pratt Trace Analysis Laboratory, and has, it is hoped, been kept general enough in its design that it may be used by any laboratory or individual in analytical chemistry. It has been made by the Charles R. Hadley Company on cards 4 inches by 6 inches, a standard filing card size, using a substantial card material (ecru Index Bristol) punched with a double row of holes all around, and printed as shown in the photograph. The following is a description of the coding system and methods used with this card at the Pratt Laboratory in a file which is predomi¬ nantly concerned with the determination of minute traces of elements by colorimetric or spectrographic methods. Author's Name. — The left hand edge of the card has three fields for coding the first three letters of the senior author’s name. The first and second field are exactly as described by Cox et al. and utilize a 6- position triangular code for the first letter and a 3-position abbreviated code for the second. The code for the third letter his been abbreviated slightly to make it fit a 5-position field. Slotting, sorting, and selecting are done as described on page 7. Date. — The first two fields at the left side of the top of the card are 4-position triangular numerical codes and are used to record the year of publication. Only the last two digits of the year are used; if a century other than the twentieth is involved the position immediately to the left of the year fields is slotted shallow. Journal. — The journal in which the article appears is coded in the triangular numerical field at the lower right hand edge of the card. Each journal (or small group of journals) is assigned a number as follows: 1951 ] Punch Card System for Bibliographies 41 TABLE IV. Information to he Coded Desirability of Coding or Usefulness When Type of Code Required Coded Sortable Selectable Enough letters of au¬ thor’s name to ensure good sorting and rea¬ sonably small dupli¬ cation on selection Highest Required Required Author’s initials Not very useful Required Required Junior authors Not very useful Required Required Complete identifica¬ tion of journal (for at least all promin¬ ent journals) Very high Not necessary but useful Required Year of publication Very high Required Required Volume, page, month, etc. Very low Element sought High Required Material analyzed Fairly high Required Method of analysis High Required Reagents High for some types of methods Not necessary but useful Required Apparatus Low Results Very low generally Required Range over which method applies Required Precision and ac¬ curacy Low Required 0) Analytical Chemistry, formerly Ind. Eng. Chem., Anal. Ed. 1 ) Analyst 2) Zeitschrift fur Analytische Chemie 3) Journal Assoc. Official Agri. Chem. 4) Journal of the American Chemical Society 5) Journal of the Society of Chemical Industry, Chemistry and Industry, and Journal of the Chemical Society 42 The Virginia Journal of Science [January Figure 14 •T? • • "1 • •> • • AUTHOR 41*1949X9^ Yoe, J. H. and A. B. Armstrong ^ " REFERENCE SUBJECT Anal.Chem. 19, 100-2 (1947) uolorimetric determination oT titanium with disodium- abstract 1 , 2-di hy d roxyben zen e- 3 , 5-di sul i on a te . "he reagent used by Yoe and Jones [C.A. ?j3,692] for Fe can used to detect 0.01 p.p.m. Ti+4 when a 50-ml.Nessler Niube is used. Interference by Fe+3 can be prevented by reducing it to Fe+2 with Na2S204 in solns. buffered at pH 4.7. The absorbancy caused by the blue Fe+3 complex can be measured spectrophotometrically at 560 mu and that 6) Spectrochimica Acta 7) Journal of the Optical Society of America 8) Mikrochemie and Mikrochimica Acta 9) Journal of Research of the National Bureau of Standards 10) Nature., Soil Science, and horticultural journals in general 11) Journal of Industrial Hygiene and Toxicology, Biochemical Journal, Journal of Biology, and biological journals in general 12) Angewandte Chemie 13) Comptes rendus 14) Bulletin de la societe chimique de France and Bulletin de la societe chimique Beiges 15) Russian journals 16) Japanese journals 17) Austrialian and New Zealand journals 18) Analytica Chimica Acta and Annales de Chimie Analytique 19) Miscellaneous Elements Sought. — The right hand portion of the upper edge of the card with holes numbered from 1-24, is used for a direct coding of elements, the following elements being given the numbers shown by symbol: 1 As 3 Bi 5 Cl 7 Cr 9 Fe 11 Mg 13 Ni 15 Sb 17 Ti 19 V 21 U 23 Others 2 A1 4 B 6 Ca 8 Co 10 Cu 12 Mn 14 Mo 16 Pb 18 Sn 20 W 22 Zn 24 No Element 1951] Punch Card System for Bibliographies 43 In order to code an element a slot is made down to the hole whose number corresponds to that element, whether it be shallow or deep. Analytical Method. — The same direct method of coding is used for the analytical method involved using the upper portion of the right hand edge of the card (holes numbered 25-40). The methods are assigned to the following holes : 25 Flame photometry 26 Spectrography 27 Electrometry other than Polarography 28 Polarography 29 Solvent extraction 30 Distillation 31 Trace gathering by co-precipitation, organic reagents, etc. 32 Ashing and Combustion 33 Gravimetry 34 Radioactivity 35 Ion Exchange 36 Chromatography 37 Treatment of data 38 Volumetry 39 Nephelometry, Turbidimetry, or Fluorometry 40 Colorimetry Material Analysed.— Along the bottom edge of the card, in holes numbered 41-54, and 62 the material being analysed, and some other general information is coded, as follows: 41 Foods and Nutritional Work 43 Photographic and Densitometric Work1 Qualitative or Spot Tests2 45 Excitation Sources1 Visual Quantitative Tests2 47 Spectrograph Design1 Instrumental Methods2 Amperometric Titration8 49 Electrodes1 Mercury Electrodes3 51 Sample Preparation1 Solid Micro-electrode3 53 Compounds — Organic 55 57 59 61 The remaining positions and fi able for future needs, as yet unfor< 42 Metals 44 Solutions 46 Minerals and Soils 48 Biological Fluids 50 Animal Tissue 52 Plant Tissues 54 Compounds — Inorganic 56 58 60 62 Punched cards dds are not yet in use and are avail- 44 The Virginia Journal of Science [January In either of the direct codes, that is, for methods or elements, it is possible to select all of the cards of one group by one of two methods. If the method or element is assigned a hole in the back row, then passing a needle through this hole and lifting will drop out all cards of this group. If the method or element has been assigned a hole in the front row, passing a needle through its hole will drop out cards of that group plus those corresponding to the hole behind it and these two groups must then be separated by needling the back holes. Although this is not a direct code in its truest form it is a method of coding the important data in a true direct code comprised of the deep holes, and then coding less im¬ portant divisions in the shallow holes where they are slightly harder to reach and may sometimes be masked by a. deep slot. If a mutually exclusive pair of data can be coded in each position the code is better, as the masking, resulting from the requirement for slotting one position both shallow and deep on one card, cannot occur. General. — On the face of the card is indicated the position for typing or writing the author’s name, the location of the reference, the subject or title of the article, its Chemical Abstracts number, and an abstract of it. Since the codes are in brown ink it has been found that typing or writing over them is quite legible and does not interfere with their use; however enough room is available on the front and back of the card that this is seldom necessary. Fig. 14 shows one of the cards with data on it, and Fig. 15 shows a “key”card, with all of the information necessary for coding written on its face. Several of these key cards are kept with the file as a handy reference when coding. The journals, elements, methods, and other categories listed above are those found to be important to the Pratt Trace Analysis Laboratory. Another laboratory might, of course, make a list quite different in its emphasis, reflecting the interests of the user. Since all the space on the card except the author and year codes can easily be adapted to almost any desired field of research, and since these two are generally necessary anyway, it is felt that this particular card may be found useful by other laboratories, especially in analytical chemistry. They may be purchased at a very reasonable price from The Charles R. Pladley Company. Future^ of such a System. — In the year and a half that this system has been used the file has shown a constant growth and with this growth it has become more and more useful. It is hoped that in time it will cover our field of chemistry as exhaustively as could a bibliography and through its extremely fast reference possibilities will provide an immense saving of time to the succeeding generations of students and researchers who pass through a laboratory of this type. 1. These designations apply only when holes No. 25 or 26 are slotted (Spectrographic or Flame Photometric) 2. These designations hold only when hole No. 40 is slotted (Colorimetric) 3. These designations hold only when hole No. 28 is slotted (Polarographic) 1951 ] Punch Card System for Bibliographies 45 Figure 15 (1) Bush, V. 1945, — Atlantic Monthly , 178: 101-8. (2) Bailey, C. F., R. S. Casey, and G. J. Cox 1946. — Science , 104: 181. (3) Cox, G. J., C. F. Bailey, and R. S. Casey 1945. — Chem. Eng. New, s. (4) Casey, R. S., C. F. Bailey, and G. J. Cox 1946. — J. Chem. Ed., 23: 495-9. (5) Cox, C. J., R. S. Casey, and C. F. Bailey 1947. — J. Chem. Ed., 24: 65. (6) Assoc, of Spec. Lihr. and Inf. Bur. Proc. 1947. — pp. 55-7. (7) Paper presented before the Chemical Literature Group of the Division of Chemical Education at the 112th meeting of the Ameri¬ can Chemical Society, New York, September 1947. (8) As 7. (9) Paper presented before the Division of Chemical Literature at 115th meeting of the American Chemical Society, San Francisco, March 1949. (10) Patterson, G. D. Jr., and M. G. Mellon 1949. — J. Chem. Ed. 26, 468. 46 The Virginia Journal of Science [January ACIDITY-ALKALINITY IN THE ALIMENTARY CANAL OF TWENTY INSECT SPECIES J. M. Grayson, Virginia Agricultural Experiment Station INTRODUCTION The importance of pH as a factor affecting the entrance of toxic substances into the tissues or body fluids of insects from the various digestive organs has been discussed in some detail by Hoskins (1940, p. 338) and pointed out by Waterhouse (1940, p. 8). Admittedly, most of the newer chemical insecticides are unaffected by the pH of the liquid in which they occur; but with many of the older insecticides, such as the arsenicals and nicotine compounds, their solubility is closely correlated with pH. Inasmuch as the stomach, or midgut, is the region of the insect digestive tract in which secretion of digestive juices and absorption of food products is thought usually to occur, the acidity-alkalinity of this region should be of special interest in toxicological work with stomach poison types of chemical insecticides. Reported here are results of the preliminary phase of a study de¬ signed to determine whether there is a positive correlation between the toxicity of certain stomach poison insecticides and their solubilities at the pH existing in the insect midgut. For example, a poison showing maximum solubility at pH 9.0 would be expected to be more toxic to those insects having a strongly alkaline reaction in the midgut than to those in which the midgut reaction was acid. If any consistent relation¬ ship, either negative or positive, could be established between the theoreti¬ cal and actual toxicities of certain stomach poison insecticides, then could not this knowledge be applied in a scientific approach towards finding new chemicals of insecticidal value? The desirability of having exact, detailed data on the acidity- alkalinity occurring in the various regions of the insect digestive tract, especially the stomach, before beginning a toxicity study of this kind should be readily apparent. Although considerable data have been pub¬ lished on the pH of the insect digestive tract, in most cases single de¬ terminations were made on the contents of whole regions, and often the data were of uncertain reliability because of the use of faulty techniques. The purposes of this study w^ere to obtain detailed information on the 1951'] Acidity-Alkalinity in Alimentary Canal of Insects 47 pH in the various regions of the insect alimentary canal; to employ a glass-electrode technique for the determination of such values; and to obtain pH data on species of insects which have not been studied pre¬ viously. REVIEW OF LITERATURE The published information on digestive tract reactions in insects of the orders Orthoptera, Lepidoptera, and Hymenoptera, are briefly summarized in Table 1. Unrecorded are the following results reported by Semans (1941) for the anterior end of the midgut in four families of Orthoptera: Acrididee, pH 6.0; Gryllacridae, pH 5.9; Tettigonidae, pH 5.6; and Gryllidae, pH 5.6. Of particular interest in the Lepidoptera are the brilliant experiments of Lnderstrom-Lang and Duspiva (1935) in which it was clearly shown that the digestion of keratin in the clothes moth larva is made possible by the presence of an extremely active pro¬ teinase, and a reducing agent, and a pH of 10.0 or higher. The most acid reaction yet reported for the insect digestive tract was found in the mid-stomach of certain species of Diptera. Hobson (1931) obtained the following pH values in Lucilia sericata : crop, 7.0 to 7.7; fore-stomach, 6.5 to 7.5; mid-stomach, 3.0 to 3.5; hind-stomach, 7.5 to 8.3; and hindgut, 8.0 to 8.5. Similar results were obtained by Waterhouse (1940) for nine other species of Diptera belonging to the genera Lucilia, Calliphora, Chrysomyia, and Musca. It is true that Randall and Doodv (1934) have reported a pH of 3.0 for the junction of the stomach and hindgut of Zootermopis angusticollis (Isoptera), but the validity of this value is open to question. Conversely, other species of flies apparently have no such acid reaction in any section of the digestive tract. Values of pH ranging from 6.0 to 7.2 for the crop, 6.6 to 8.2 for the midgut, and 6.2 to 8.0 for the hindgut have been found in Psychoda and Chironomus (Crozier, 1924), in Glossina (Wiggles worth, 1929), in Hypoderma lineatum (Simmons, 1939), and in seven species of mosquitoes (Senior-White, 1926). The remaining data, relative to insect digestive tract reactions, are rather meager and pertain to a number of insect orders. Values of pH for 25 species of Coleoptera have been reported by Swingle (1930 and 1931a), Kruger (1933), and Staudenmayer and Stellwaag (1940); three species of Hemiptera by Swingle (1931a); two species of Odonata bv Swingle (1931a) and Kruger (1933); two species of Tricoptera by Kruger (1933); and a single species of Neuroptera by Swingle (1931a). Additional observations have been reported by Popow andGolzowa (1933) for two species of mosquitoes, a bed bug, a tick, and the human body louse. 48 The Virginia Journal of Science [January (NNO(TOM cdc-^E^cdc^COCOc^cd CO o cd cd I I i>cq^>-;cocoinoco idiOLOLdcd'^-riHCO'^ I I I 12 I I I I I I I I o 05 os' ’ ~ 1 CJ 05 co ^ ,05 cd cd o5 cd w hhhh H e, 22 22 22 22 ’2 5 > o ° g OJOdJCUm^^S W) tfltl) tl)p2 08 F-H ij .s.s.s.s ss ^ s 0/2 C/2 C/2 C/3 1> Q O C/2 <<< tuO tuO tuO > > > cd cd cd m c/5 xn CD CD +j ^ g O W 03 tuo tuo $ M’S Si? 6 ISg-sS|SS33 oouugoo^g CO CO q ^ 05 05 1—1 cd OOffiOO^'tCOCOlOOo 05 05 05cd05cdc3c005cdcd I I I I I I5S2S I I II I II I I I II I ! I I I I I I I I I U I I MS II II O '“■' CO CO ^ co 22 S £ 22 05 CT> 1—H 22 05 05 05 F— 1 cd cd CO CO 05 05 cd T5 o c, . cd cd cd rxj Sh i— i cd a) .p a tuo 41; w> a as3 w P W W*JH ?-< p 3 ^ Cd CD CD p fi 3^ '%'%S c/2 c/2 c/2 43 ►JhJhJ O s s s s oooS S> S> £► Jh Q hJ 43 -4-J o j •! B^'B 2^4^ in C cd CD ^ -m B x § . cnmmou>0 a> -1 1 a ,Hj w H x £.SS | 043 O ih 3 a £ d s w ^ d2 tuo cd 2^ S3 a ft P^ld Cl co cd lo cd so 05 ^t1 co I co ^ LO cd cd * c3 cd cd I I O OC CO CN! ■ 'd cd *d cd I I 1 I 1 I I. I I I. I I. S3 o ?_i cd cd cd cd ^HHHH |£d cd co co co co £3 hh 05 05 05 05 05 HH ,— i ,— | wog WWWW ^ in ^CD O) £<1 Qj ® n =a&|SS ffiWo&Wm * A=adult; L~ larva ** Fluid emitted from mouth 1951 ] Acidity- Alkalinity in Alimentary Canal of Insects 49 Method The insect specimens were collected in their natural habitat, brought to the laboratory, and pH determinations made as soon thereafter as was possible. They were placed in cages and provided with food if determina¬ tions could not be made within a few hours after collection. Each insect was dissected by clipping off any large appendages such as legs and wings, exposing the entire body cavity by a longitudinal incision through the dorsal integument, and removing any tissues preventing quick access to the. alimentary canal. Mixing of gut contents was prevented by either tying a silk thread tightly around the insect gut at appropriate places, or clamping the gut with tweezers. In obtaining a sample, a small incision of the gut wall was made with an insect pin or pointed glass rod, and the whole gut contents withdrawn by means of a micro-pipette and fine- pointed tweezers. Usually the samples obtained were moist enough to insure electrical conductivity, but sometimes it was necessary to add a trace of distilled water. This was done by first injecting the water into the insect gut at the desired point and then withdrawing it for addition to the sample. Complete determinations were made on each insect in those cases where satisfactory dissections were made and no delay in time occurred. Doubtful pH readings arising from such causes as too small a sample, eruption of the gut wall during dissection, or undue delay in completing the readings, were discarded. In most cases the insect specimens selected for study were those containing considerable food in the digestive tract. Thus larger samples were available, and the condition of the digestive tract was considered as perhaps being comparable to that following ingestion of poisons with a normal diet. The actual pH determinations were made with a Model G Beckman pH meter in combination with a calomel electrode and a one-drop, open glass electrode. Apparently there is only one reference in the literature (Duspiva, 1936) to the use of the glass electrode for determination of insect digestive tract pH, although this device has been used for pH determination of insect body fluids by Taylor and Birnie (1933), Taylor et al (1934), and Craig and Clark (1938). Duspiva (1936) has shown that small samples of high alkalinity may be affected from contact with carbon dioxide of the air, but the writer has found that high alkaline samples of the type obtained in this study are unchanged after exposure to air for as long as five minutes. Results Table 2 lists the pH values obtained in the various regions of the alimentary canal of eight caterpillar species and one species of sawfiy larva. All of the Lepidoptera larvae were feeding on plant leaves at the time of collection, except the corn earworm which was feeding on the 50 The Virginia Journal of Science [January kernels of developing corn. The sawfly larvae were feeding on the leaves of dogwood. From 10 to 18 individual readings were made for each region of the digestive tract in the case of all species except the larvae of Datana angusii where each average figure was obtained from only 8 to 9 individual pH readings. The general average pH values obtained for the eight Lepidoptera species were: fore-stom?ch, 8.4; mid-stomach, 8.5; hind-stomach, 7.7; and rectum, 6.3 (Figure 1). The pH values obtained in the various regions of the alimentary canal of ten species of adult grasshoppers and the adult stage of the American cockroach are shrown in Table 3. The grasshoppers were col¬ lected from their normal habitats and presumably had been feeding on grasses, clovers, sedges, or other similar plants. The American cock¬ roaches were taken from a laboratory stock culture in which the food provided wras dried dog biscuits, lettuce leaves, and water. Separate determinations were made on males and females in the case of four grasshopper species. Most average figures shown in Table 3 were de¬ termined from approximately 10 or more individual pH readings each; however, only 4 or 5 readings were obtained for each digestive tract region in the case of Melanoplus bivittatus. Readings on saliva were made from the liquid emitted from the mouth, which probably consisted of the secretions of the salivary glands plus regurgitated liquids from the crop. The general average pH values obtained for the ten grasshopper species were: saliva, 5.5; crop, 5.6; fore-stomach, 6.4; hind-stomach, 6.7; intestine, 6.7; and rectum, 6.2 (Figure 1). The approximate locations in the different insect digestive tracts from which the samples were taken are indicated on the sketches shown in Figure 2. Discussion The one-drop glass electrode was found to be very satisfactory for measurement of insect digestive tract pH when samples of approxi¬ mately 0.5 drop or larger were available. Rather rapid readings can be made with the aid of an assistant to dissect the insect specimens. The acidity-alkalinity found in the digestive tract of adult grass¬ hoppers as reported here is in fair agreement with the results reported by Bodine (1925) but somewhat at variance with those of Swingle (1931a). Critical comparison in either case is difficult because each author made only one reading for entire midgut and hindgut regions. Inasmuch as the results reported by Bodine are really pH values of the grasshopper gut wall, rather than gut contents, it is perhaps surprising that his data would agree so closely with those presented here. This would indicate that the ingestion of food has little effect on the acidity- alkalinity of the grasshopper digestive tract. The pH values obtained in this study for the digestive tract of the American cockroach are similar to those reported by Wigglesworth 195l~\ Acidity- Alkalinity in Alimentary Canal of Insects 51 Table 2. — Values of pH obtained in the alimentary canal of eight species of caterpillars (Lepidoptera) and one species of sawfly (Hymen- optera). _ Stomach _ Species Crop Fore Mid Hind Rectum (Lepidoptera) Ceratomia catalpae . Range — 8. 5-9. 8 8.6-10.1 7.0-8.7 5. 6-6.7 (Bdvl.) . . . . Avg. — 9.0 9.3 7.8 6.1 Protoparce . . . Range — 8.3-10.1 8.4-9. 9 7.3-8.6 6.5-7.3 quinquemaculata (Haw.) . . . . . Avg. — 9.3 9.2 7.8 6.8 Anisota senatoria . Rnage — 8.4-9.6 8. 1-9.9 7.2-8.2 5. 5-6. 7 (A. & S.) . - . . Avg. — 9.0 9.1 7.8 6.0 Hyhantria cunea . Range — 7. 6-8. 6 7. 6-8.5 7. 0-8.5 5. 5-7. 3 (Drury) . . . . . Avg. — 8,1 8.0 7.7 6.1 Datana angusii . Range — 7. 9-8.8 8.8-9. 6 7. 8-8.8 - G. & R . . . Avg. — 8 .5 9 .0 8 .3 Archips jervidana . Range — 7.5-9. 9 7.5-9. 3 7. 2-8.3 - (Clem.) . Avg, — 8.3 7.9 7.8 _ Heliothis armigera . Range — 7. 1-9.1 7. 0-9.4 6. 6-7.8 5.8-7.5 (Hbn.) . . Avg. — 7.8 8.0 7.2 6.6 Pieris rapae . Range •— 6. 8-7. 7 6. 9-8.1 6.4-7.9 6. 7-7. 8 (L.) . . | . ......Avg. — 7.3 7.6 7.4 7.2* (Hymenoptera) Macremphythus varianus . Rge. 5. 1-5.9 6.3-7. 2 6. 3-7.7 5.7-6.4 (Nort.) . I . . . Avg. 5.5 6.6 7.1 6.1 * Determinations made on freshly voided faeces rather than rectal contents. (1927) and Day and Powning (1949) for the German and American cockroaches. It is of interest that they found the cockroaches to have a pH of 6.0 to 6.3 in the crop after being fed on a strictly protein diet but a pH of 4.5 to 4.8 after they had fed on a carbohydrate diet. Apparently only two of the Lepidoptera species reported in this study have been included in previous work on digestive tract aciditv- alkalinity. The pH values obtained by Swingle (1931a) for the storm oh contents of the catalpa sphinx and the corn earworm agree rather closely with those reported here. Also, the reaction obtained in the gut of Pieris brassicae by Staudenmayer and Stellwaag (1940) is in fair agreement 52 The Virginia Journal of Science [January 05 Oi o X T3 C3 05 Ph £ 05 .£ H-> cn CL ■s o ffi o ° 3 3 2 3 Ad k5 O > ® e g CO ^ CO l> r— 1 05 CO CO CO o m «?© ^05 *7 CO ^ i* *7 n. i 05 P3 co cd ^ cd co id cm cd 05 CO t-I CO lO CO io id o id id id LO LO id cd id LO LO 05 Pi H LO l> CO ^ co 05 lO co 05 t-H C/5 CD 'J cq ‘J Ld CO h-+I 1 ^ *7 <=2 05 *>C0 ‘JiO c7® ‘Jo cd ^ i ^ tfco CO CO "H CO H CO CO o CO CO O CO 05 cd LO CO ^ C-3 O CO lo cd Pi 1— 1 id cd cd id cd cd LO cd cd cd id o CO LO CO t-h cm CO LO lA t-H lO T3 *7 «? J CO ‘J n *>.> i*-. ! ^7 '-i J CO *Jcq (H 05 CO CM CO 05 CO 05 CO CO CO 00 CO CO CO CO CM cd CM cd o cd Ah a s lo cd LO LO id id id cd cd cd cd CO Ci PI o 00 CO H 05 LO C- LO LO H-> m 05 Ph L7 in i<=>. i*. — i cd t- CO 05 CO co cd c— cd lA CO 05 CO CO CO O co 05 CO lo cd LO id id id id id cd cd id O 05 L'- CO 05 co o 05 05 05 o s_. ^LO ^LO i«=>. ^ __LJ lO 1 ■ 1 ^ 1 id i ^co id _ 1 id ^ i o 00 LO TjH LO CO LO CO LO O LO CM LO CO lo CM LO ^ LO CO LO LO LO id id id LO id id id LO id id id CO I> o LO 'cH t> i> o CO L— 00 > l? ^ °? CO LO 1 CO cd | LO id id ^ i co | cd _ i ^ ^LO “?>o ^co 73 i lo co id Pi £0 Pi w fl 5® Pi ^ fl W) P! cd > cd > cd > cd ^ cd cd ^ cd > cd > cd > cd > cd > X &< Ph< P3< K5 Ph /-s r"N 43 P5 >> cd Ph CD H g Ph Q a 0) C/5 rd PS o in CO m CO S3 -+H> cd 3 £ co co g S 05 S p. rO S P- 1 P- N-/ cd e cd CJ *'S Ph . d cq cd P. Ph O § g S3 3 h-h IS P. 05 05 e CO CO O 05 •»* ’B IS cd P- "§ <05 s ^5 g cd a o Ai g a 05 C j P. CO a co CO cd CJ ss s O 05 a CO S3 g a cd g a g a p. 05 CJ cd H g s 3 £ S H-i ■c-A Cd a g rO g g CO p. a •<-H US ^5 P- a in "S IS CJ E- cd § 1951] Acidity-Alkalinity in Alimentary Canal of Insects 53 o CO OS © * CO * *7 CO ^CO cd * i>! * I CD | LO LO CO NlO ^ cd T_! I> ® L-- lo id id r- d- CD CO LO t-H *>!> ^7 CD *7 CO | t-H | IV i-H cd co cd ^ cd CO^ CO zo cd lo cd cd cd LO co ^ tv iv I ^7 lo ^7 os CD CD 1 Cd 1 Cd CH] CO oo cd ^ cd co cd lo cd cd id cd id lo CO 1 I> 1 Dd | CO LO ^ lo ^ id ^ id id LO id id Tf ^ o CO © cd lo cd 1 CD 1 CD I CD ^ LO id od id LO id id 0) 0 0) 0 0 tuO • W> • tuo . tuo • tuo . g W) c| bJO rj bfl o tuo O tuo cd > cd > cd > cd > cd > Ph< P$> CJ C/j ►4 cd • cd N-'' S Vs 'S cd g C3 CJ j- 0 • c-i o £ i- g cd o Cj cd cd cd V> S— •■«» 0 a CO S o +n cd CO O P« p &0 CO • s- 0 o 5 0s £5 i * O o So S 8S •Th cd BB *S 2 o 3 -M ,Z