Reflections on Experimental Science

The following sections are included:

• CONTENTS

• PREFACE

The following sections are included:

• Introduction

• SLAC, Leptons, and Heavy Leptons

• Heavy Lepton Searches in the 1960's

• Photoproduction Searches for New Charged Leptons

• Studies of Muon–Proton Inelastic Scattering

• Electron–Positron Colliding Beams and Sequential Leptons

• The SLAC-LBL Proposal

• Lepton Searches at ADONE

• Discovery of the Tau in the Mark I Experiment: 1974–76

• The 1975 Canadian Talks

• First Publication: “We have no conventional explanation for these events”

• Is it a Lepton? 1976–1978

• Anomalous Muon Events

• Anomalous Electron Events

• Semileptonic Decay Modes and the Search for τ⁻→ν_τπ⁻ and τ⁻→ν_τρ⁻

• The Tau Mass

• The Tau Lifetime

• References

A search for new particles which might be produced by photons of energy up to 18 GeV is described. No new particles were found. Calculations of the Bethe-Heitler process are described which make it possible to state that this experiment would have detected non-strongly-interacting particles whose mass and lifetime lay in a definite range, did they exist.

As a test of muon-electron universality we have compared muon-proton and electron-proton inelastic scattering cross sections for |q²| (square of the four-momentum transferred from the lepton) values up to 4.0 (GeV/c)² and for lepton energy losses up to 9 GeV. There is no experimentally significant deviation from muon-electron universality. If the muon is assigned the form factor relative to the electron , then with 97.7% confidence Λ_d > 4.1 GeV/c.

The following sections are included:

• I. Introduction

• II. Some Possible Types of Heavy Leptons
  • Heavy Sequential Leptons: μ^', μ^" …
  • Heavy Excited Leptons: e^*, μ^*
  • Special Pairs of Heavy Leptons
  • Stable and Very Long-Lived Heavy Leptons
  • Other Possibilities
  • A Word on Notation

• III. The Decay Properties of the Heavy Leptons
  • Charged Heavy Sequential Leptons
  • Neutral Heavy Sequential Leptons
  • Special Pairs of Leptons
  • Charged Heavy Excited Leptons

• IV. Searches for Low Mass or Stable Heavy Leptons
  • Searches in the Decay Modes of the Pion and Kaon
    • The Method
    • Past Searches
    • Future Searches
  • Searches in Particle Beams for Short-Lived Charged Heavy Leptons
    • The Method
    • Past Searches
    • Future Searches
  • Searches for Stable Charged Heavy Leptons in Particle Beams
    • The Method
    • Past Searches
    • Future Searches

• V. Searches for Unstable Heavy Leptons with Masses Greater Than About 0.5 GeV/c²
  • Detection Method
  • Electron-Positron Colliding Beams Production of Heavy Leptons
    • The Method
    • Past Searches
    • Present and Future Searches
  • Photoproduction of Heavy Leptons
    • The Method
    • Past Searches
    • Future Searches
  • Heavy Lepton Production in Proton-Proton Collisions
  • Searches for Heavy Excited Leptons Using Charged Lepton-Proton Eleastic Scattering
    • The Method
    • Past Searches
    • Future Searches
  • Searches for Heavy Excited Leptons in Lepton Bremsstrahlung
    • The Method
    • Past Searches
    • Future Searches
  • Searches for Heavy Leptons Using Charged Lepton-Proton Inelastic Scattering
  • Searches for Special Pair Heavy Leptons Using Neutrino-Nucleon Inelastic Scattering

• VI. Comparison of Some Static Properties of the Muon and the Electron
  • Electric Charge
  • Gyromagnetic Ratio

• VII. Mu-Mesic Atoms

• VIII. High Energy Reactions of Mons and Electrons

• IX. Muon-Proton Elastic Scattering and Form Factors
  • Theoretical Background
  • Experimental Results

• X. Muon-Proton Inelastic Scattering
  • Theoretical Background
  • Experimental Results

• XI. Charged Lepton Form Factors in Colliding Beam Experiments
  • Elastic Electron-Electron Scattering: e⁻+e⁻→e⁻+e⁻
  • Bhabha Scattering: e⁻+e⁺→e⁻+e⁺
  • Muon Pair Production: e⁻+e⁺→μ⁻+μ⁺

• XIII. Speculations
  • Searches for Heavy Leptons
  • Charged-Lepton Form Factors
  • Anomalous Lepton-Hadron Interactions

• References

The following sections are included:

• Introduction
  • Heavy Leptons
  • Heavy Mesons
  • Intermediate Boson
  • Other Elementary Bosons
  • Other Interpretations

• Experimental Method

• Search Method and Event Selection
  • The 4.8 GeV Sample
  • Event Selection

• Backgrounds
  • External Determination
  • Internal Determination

• Properites of eμ Events

• Cross Sections of eμ Events

• Hypothesis Tests and Remarks
  • Momenta Spectra
  • θ_coll Distribution
  • Cross Sections and Decay Ratios

• Compatibility of ee and μe Events

• Conclusions

We have found events of the form e⁺+e⁻→e^±+μ^∓+ missing energy, in which no other charged particles or photons are detected. Most of these events are detected at or above a center-of-mass energy of 4 GeV. The missing-energy and missing-momentum spectra require that at least two additional particles be produced in each event. We have no conventional explanation for these events.

The existing data on e^±μ^∓, e^±x^∓, μ^±x^∓, and related events produced in e⁺e⁻ annihilation are reviewed. All data are consistent with the existence of a new charged lepton, τ^±, of mass 1.9 ± .1 GeV/c².

The following sections are included:

• Introduction

• τ Production and Related τ Properties
  • e⁺+e⁻→τ⁺+τ⁻
    • Threshold Region and m_τ
    • Above Threshold to About 10 GeV
    • Above 10 GeV to Below Z⁰ Resonance
    • Z⁰ Resonance
    • Above the Z⁰ Resonance
    • Some τ Studies Related to e⁺+e⁻→τ⁺+τ⁻
  • Photoproduction: γ+N→τ⁺+τ⁻+N′
  • Particle Decays to τ and ν_τ
    • W⁺→τ⁺+ν_τ
    • D Decays to τ and ν_τ
    • B Decays to τ and ν_τ
  • τ Production in Hadron Collisions
    • τ Production in p + N Collisions
    • τ⁺τ⁻ Production in Heavy Ion Collisions

• General Discussion of τ Decays
  • Overview of τ Decay
  • Overview of Branching Fractions
  • Topological Branching Fractions
  • Unconventional Decays
    • No-ν_τ
    • With X⁰
    • Non-W Exchange

• Leptonic Decays
  • Overview of Decay Widths and Branching Fractions
  • Crude Calculation of B_e, B_μ, and B_had
  • Precise Calculation of Γ_e and Γ_μ
  • Aside on Radiative Decays
  • Comparison of B_e, B_μ, and T_τ Measurements
  • Momentum Spectrum in Leptonic Decays

• Hadronic Decays
  • τ⁻→π⁻+ν_τ, τ⁻→K⁻+ν_τ
  • Application of Quantum Number Conservation in Non-Strange Hadronic Decays
  • τ⁻ → ρ⁻ + ν_τ and Other Vector Decay Modes
  • Measurements of Modes Containing K Mesons
  • Comparison of B₁ and B₃ with Σ_i B_i
  • Measurements of Strong Interaction Constant α_s

• Tau Spin Phenomena
  • τ Spin Alignment and Decay Correlation
  Polarization at the Z⁰
  • Search for CP Violation in τ Production
  • Search for CP Violation in τ Decay
  • Tau Magnetic Moment

• The τ Neutrino ν_τ
  • ν_τ Mass Limits
  • ν_τ as Dark Matter
  • ν_τ Lifetime Limits
  • ν_τ Weak Interactions
  • The ν_τ and Neutrino Mixing
  • ν_τ-Nucleon Interactions

I wrote this paper in 1991 about three years after Jasper Kirkby and John Jowett conceived the idea of a tau-charm factory, a high luminosity electron-positron collider which would operate in the threshold region for the production of tau lepton pairs and charm meson pairs. In 1995 there is no tau-charm factory built or even under construction, although there is a strong interest in such a research instrument at the Institute for High Energy Physics in Beijing.

Beginning in 1988, Juan Antonio Rubio, Jasper Kirkby, John Jowett, Rafe Schindler, I, and many other tried hard to get a tau-charm factory built. We carried out studies of the potential physics, we designed the collider and the detector, we write proposals, we lobbied European and United States high energy physics laboratories, we lobbied funding agencies in Europe and the United States, we organized workshops and issued proceedings, we tried to get a tau-charm factory built at CERN in Geneva, at SLAC in the San Francisco Bay Area, in the south of Spain near Seville. Other physicists considered sites in Russia and in France…

A tremendous amount has been learned about quantum mechanics, quantum electrodynamics, and elementary particles by the experimental study of atoms, that is electronnucleus systems, and the experimental study of positronium, the electron-positron atom e⁺e^-. Similarly a great deal has been learned from mu mesic atoms, the system consisting of a negative muon and a nucleus usually with additional electrons.

This paper concentrates mainly on the properties of systems in which a tau lepton replaces an electron or positron, namely an atom containing a negative tau lepton or a system consisting of a negative tau lepton and a positive tau lepton τ⁺τ^-. These thoughts came out of my work on the tau-charm factory because that is the only instrument where we might hope to produce such tau atomic systems.

But can we learn anything new from tau atomic systems even if we could make them and study them? I have not been able to propose clear research objectives. But in physics studying a system for it's own sake can be pleasurable and there is always the dream of the unexpected.

As of Spring, 1995 this was my most recent paper on tau lepton physics. There has been a tremendous amount of experimental research on the properties of the tau since my discovery of the tau in 1975. There have been a thousand experimental and theoretical papers and probably a hundred Ph.D. theses. Research techniques have become more and more refined and there has been a steady increase in the amount of data on tau production and decay from experiments at electron-positron colliders.

There are two summarizing observations to be made about our present knowledge of the physics of the tau lepton. First, contrary to earlier hopes that the tau possessed some non-observed decay modes (the one-prong problem) all existing measurements confirm that the tau obeys conventional theories of weak and electromagnetic interactions. Second, as the amount of data increases and statistical measurement errors decrease in the study of the tau decay modes, measurement uncertainties are becoming dominated by systematic errors.

In this paper I try to project how much these systematic errors might decrease in the future and this illustrates a general problem in projecting scientific research. I can guess, sometimes quite well, how much present particle detector technology will improve. But I cannot guess what new technology might be applied to the study of the tau. If I had an idea for a new and powerful technology for studying the tau or other elementary particles, I would be writing papers about it and looking for funds to realize it. I would do this in spite of my discouragement about the time I spent on the tau-charm factory…

This atomic beam experiment was my first physics experiment, my Ph.D. thesis research and my first physics paper. Now, forty years later, I remember little about the significance of the measurement itself, the determination of the electric quadrupole moment of the sodium nucleus, but I remember a great deal about building and running the atomic beam apparatus.

I was lucky that the general nature of this experiment coincided with the aspects of scientific work which gave me the most pleasure. First of all the experiment was grandly mechanical and physical: the large brass vacuum chamber; the vacuum pumps, the physical beam of sodium atoms; the mechanical adjustments for positioning the magnets, lenses, and detector along the atomic beam. Even the power supply for the magnets, submarine batteries charged with a motor-generator set, was more mechanical than electronic.

In most of my later experiments I have preferred the mechanical to the electronic; I love the mechanical, the electronic I tolerate. It is true that mechanical has taken on for me very broad meanings. Thus for me optics is a mechanical subject and my ways of thinking about fundamental physical processes are also mechanical. In this first experiment I saw in my mind the photons hitting the sodium atom in the beam, knocking them up to an excited state, and then the radio frequency electromagnetic field twisting the excited atoms into a different hyperfine structure state. It has been the same in my work on the tau lepton. I see elementary particles as very small mechanical objects which collide and decay…

After receiving my PhD. in atomic physics at Columbia University in 1955 I went to the University of Michigan to work in elementary particle physics. My changing physics fields was mostly due to the urging of my thesis professor, Isadore Rabi. Although he never worked in elementary particle physics, he saw its great potential; also, he liked to annoy his fellow atomic physicsts at Columbia by telling his students that atomic physics was a dying field.

I chose Michigan because Donald Glaser had just invented the bubble chamber at Michigan. For about the next ten years I worked intermittently in bubble chamber experiments. I enjoyed the first few experiments where we used a non-magnetic propane bubble chamber with a small sensitive volume of 12 cm by 12 cm by 30 cm. We built the entire chamber in the Physics Department of the University of Michgan and the construction involved interesting mechanical and optical work The entire apparatus was small enough so that we could push it into and out of the particle beams of the Cosmotron at Brookhaven.

Runs lasted a few days to a week. I remember that as I would come on shift walking onto the floor of the Cosmotron, I would listen for the periodic expansion noise from the chamber. The same kind of run excitement as I felt during my atomic beam experiment. Was the chamber still working? Would the expansion system last the run?

This reprint describes one of our first experiments using the propane chamber. It is interesting how in the first section we argue the advantages of a bubble chamber over an emulsion. This was the beginning of my interest in the elastic scattering of elementary particles, I thought that this simplest of reactions could reveal the most about the interaction of particles…

Although busy with bubble chamber physics in the late 1950's I kept looking for a new particle physics technology which would combine the best features of bubble chambers and counters. The magnetic bubble chamber technology provided large solid angles, precise measurements of vector momenta, good vertex identification, dE/dx information, and sometimes photon detection. But bubble chambers could not be triggered at a high rate on selected events, and the scanning and measuring of bubble chamber pictures had become an industry. On the other hand counters could be used to study selected events at a high rate, but they provided limited information.

When my friend and colleague Lawrence Jones heard about a new idea in the Soviet Union, the luminescent chamber, he and I immediately started to build one. Part of our motivation was to be on our own in particle research and so we wrote to the Office of Naval Research for research support. We were young and unknown, instructors or assistant professors, I don't remember which, but SPUTNIK had just gone up and Washington was worried about a Russian lead in science. What better scientists to support than two young Americans, who wanted to work on a new Russian idea in high energy physics.

Figure 1 in Reprint C3 shows the idea: a visible light image intensifying tube with sufficient amplification allows photography of the trail of a charged particle moving through a scintillator. The volume of scintillator acts as a bubble chamber, the time discrimination of the image intensifier allows high rate triggering on selected events.

We had two problems: how to balance high light collection efficiency against large depth of field, and how to obtain enough light amplification. To ameliorate the first problem we used thallium-doped sodium iodide viewed through f / 5 aperture lenses. The second problem occurred because there were no image intensifier tubes of sufficient amplification available to us. Most image intensifier tubes were being made for military use and were classified. We had to cascade three tubes to get enough amplification…

Please refer to full text.

I loved the technology of optical spark chamber experiments: the mechanical problems of building a large, rugged, precise chamber; the high speed cameras; the optics and optical paths that extended over meters; the film scanning and measuring tables. For more than a decade I used optical spark chambers to explore pion-proton elastic scattering, neutronproton elastic scattering, muon-proton inelastic scattering, and finally electron-proton inelastic scattering.

For me the combination of optical spark chamber technology with elastic scattering experiment was particularly satisfying. Lawrence Jones and I were entranced by the early 1960's interest in the Regge pole model of strong interactions, and the cleanest example: elastic scattering. Bubble chamber experiments could not provide enough data and counter experiment only worked with high intensity proton beams. The optical spark chamber was ideal.

The paper reprinted here describes one of several pion-proton elastic scattering studies we carried out at the Bevatron. We had enough statistics to study two burning issues in the Regge pole model of the 1960's. As the energy increases, does the diffraction peak become narrower or, to use the jargon of the period, does the diffraction peak shrink? The second burning issue was the existence and significance of secondary maxima in the elastic differential cross section…

This middle 1960's experiment was designed in an afternoon. I had already moved from the University of Michigan to the Stanford Linear Accelerator Center. Michael Longo from Michigan and I had asked the Bevatron Program Committee for time to study very small angle π^– – p elastic scattering in the angular region when coulomb scattering and diffractive scattering interfere.

On that afternoon we were talking about this proposed experiment but we were dissatisfied with it. I don't remember the precise sources of our dissatisfaction. I think we were worried that our experimental design was faulty and we were not sure we could reach the accuracy we wanted. We began to look for another idea and realized that we could do an accurate and broad study of high energy neutron-proton elastic scattering using a new technique.

There were no previous measurements of high energy n – p scattering because no one knew how to make a single energy neutron beam. We didn't know how to make one either, but we realized we didn't need one! We proposed to use an incident neutron beam of all energies, scattering on a hydrogen target. We would measure the angles and momentum of the scattered proton and we would measure the angles and energy of the scattered proton and we would measure the angles and energy of the scattered neutron. The scattered neutron energy would be obtained from the interaction of the neutron and a set of thick plate optical spark chambers, an early form of a hadron calorimeter. We went before the Bevatron Program Committee and they accepted our new proposal. Proposing experiments was simpler in those days.

The experiment was led by Michael Kreisler as his Stanford PhD. thesis. As you can see from the author list there were just five of us on the experiment, but now we had help from four technicians including, once again, Orman Haas from Michigan…

In Part A of this book I described my work on muon-proton inelastic scattering. This search for differences between the muon and the electron, by its failure, led to my discovery of the tau lepton. Here I have reprinted the article describing the muon beam itself. It was a neat design.

A 17 GeV primary electron beam from the SLAC linear accelerator hit a water-cooled copper target producing muon pairs via bremsstrahlung and pair production. The target was immediately followed by a 5.5 m long beryllium filter to remove pions. The filter was made of 30 cm long sections, each section had an 18 cm diameter. This was, and still is, a lot of beryllium. We obtained it from the Rocky Flats National Laboratory and after the muon experiments were done, we lent it to other high-energy laboratories.

The size and complexity of the muon beam made these muon-proton experiments my first big engineering experiment. It was a taste of the even bigger engineering jobs that would be required to build electron-positron annihilation detectors. Since my first career love was engineering, this was a pleasant return to earlier interests. More than pleasant, there is a great thrill in bringing a large apparatus from the first sketches through to construction and operation.

In the middle and late 1960's my research concentrated more and more on the nature of leptons and the muon-electron problem. But this research itself led us back into the hadron world in the following way.

My primary interest in muon-proton inelastic scattering was to search for muon electron differences as described in Memoir A1 and Reprint A3. In the muon-proton inelastic scattering experiments we simply measured the angles and momentum of the scattered muon, ignoring the hadrons, mostly pions, that were produced. These hadrons were produced by a virtual photon passing from the muon to the proton. Again my mechanical view of particle physics. We wondered how the degree of virtuality of the photon would affect production of hadrons. How would the multiplicities and kinematic distributions depend on q², the square of the four-momentum transferred to the proton?

The muon beam had a large diameter which made it difficult to detect the produced hadrons. Also we wanted higher incident beam energy. The ideal experiment would use a high energy, small diameter electron beam. And so we designed and built an experiment to measure the scattered e^– and forward produced π's and K's in

e^– + p → e^–+ forward π's and K's + other hadrons…

There is a fifteen year gap between the time we did the research described in Reprint C8 and the period of this paper. Almost all my research in that fifteen year period was immersed in the discovery and exploration of the tau lepton; described in Part A of this volume.

This reprint raises the question of why should any individual experimenter labor on an experiment, why not wait until someone else does the experiment? Let me explain. In the mid-1980's two Z⁰ facilities were under construction, the LEP circular electronpositron collider at CERN and the SLC linear electron-positron collider at SLAC. During the proposal and construction of these colliders and the concurrent construction of the needed detectors, the involved high energy physicists organized extensive study groups and workshops on the potential physics of the Z⁰. Since then, not only the construction, but even the proposal of a large high energy facility, has brought forth extensive workshops on the potential physics, workshop with voluminous proceedings.

As I wrote in the comment on Reprint C2,1 don't do well in specialized research areas where there are already enough experimenters at work. The potential for new physics discoveries in the study of the Z⁰ attracted many able experimenters, I felt not needed. I wonder how many other elementary particle experimenters felt that way or feel that way today.

And I wonder how many experimenters feel as I do that there are an excesive number of workshops on the physics which may be done at the large accelerators either proposed or under construction. Of course, substantial thought and planning must be devoted to the design of new colliders and detectors. But do other high energy physicists sometimes think, as I do, that there is no need for endless Monte Carlo studies of signals for new particles. Isn't the actual carrying out of a new particle search the best stimulus for new ideas about how to make the search…

the question of whether quarks could exist in isolation, as separate particles with electric charge ±1/3 or ±2/3. Of course, by the middle 1980's this was already an old question and there had been many searches for isolated quarks. We were not impelled to begin a search for isolated quarks by new theoretical ideas of ourselves or others. Rather we thought up a new method to do the search. Our original idea was to move a small piece of matter periodically back and forth between two Faraday cups, thus producing an oscillating current proportional to the net electric charge on the piece of matter.

As described in this paper, it turned out to be better to leave the piece of matter at rest and to move the Faraday cups past the piece of matter using a rotating cylinder supporting a ring of Faraday cups. We were able to measure the charge with a resolution of 0.3 of an electron charge. We needed better accuracy, about 0.05 of an electron charge, and we believe we could have achieved that accuracy.

But we ran into a problem we could not solve. We had to be able to move the piece of matter away from the Faraday cups without the charge changing on the piece of matter. This was necessary so that the actual size of the charge could be measured. Our plan was to have a small hollow cylinder, a sample holder, always close to the Faraday cups and to move the piece of matter into and out of the sample holder. As described in the paper, we could not see how to do this.

When I think back to this experiment, I ruminate as to what lessons this experiment contains for the experimenter. Should the scientists always have a complete plan for an experiment with all problems solved before the experiment is built? That's too stiff a rule because it is often necessary to be working physically on an apparatus in order to solve a problem with that apparatus.

Our rotor electrometer experiment was not continued because there were actually two problems: how to move the piece of matter in and out of the sample holder without changing the charge, and how to improve the charge accuracy to 0.05 of an electron charge. Solving both problems was too much work. So the rule is, avoid experiments which have two serious problems.

I returned to the search for isolated quarks last year with a new version of the old Millikan liquid drop method; this is described in Reprint C14.

In 1988 I reviewed all of electron-positron collision physics for a meeting of nuclear and particle physicists. While preparing this review I realized that there was no experimental data on e⁺e^– collisions in the center-of-mass energy region between several MeV and about 400 MeV. There still is no data in this region in 1995. At any energy the reactions

e⁺ + e^– → e⁺ + e^–

e⁺ + e^– → γ + γ

will occur. Above 200 MeV

e⁺ + e^– → γ⁺ + γ^–

ee⁺ + e^– → μ⁺ + μ^– + γ

will also occur

This research has not progressed beyond this Physical Review paper by Christopher Hawkins and myself. Like the study of electron-positron collisions in the several MeV to several hundred MeV energy range discussed in Reprint C11, this study was stopped by my work on the tau-charm factory. However now that I am working on electron-positron physics with the CLEO experiment I will be able to take up this study again.

The speculation is that just as the strong force only involves quarks and gluons, there might be a force which only involves charged leptons, a charged-lepton-specific force. My hope is that this force would explain the large, perhaps infinite, ratio of the mass of a charged lepton to the mass of its associated neutrino.

When I began to think about searching for this force using data from the CLEO experiment, I realized again how important it is to me, and I think for most expermental scientists, to be carrying out experiments in order to get new ideas. Some colleagues in the CLEO experiment have already studied the rare decay modes of the tau…

This was a rare experiment for the 1990's world of high energy physics: it was built in six months, mostly out of parts of previous experiments; data was acquired in a month; and a handful of physicists carried it all out. In the 1950's the Russian physicists Landau and Pomeranchuk predicted that the rate of emission of photons by ultra relativistic electrons would be affected in a peculiar way by the density of matter. The prediction was made more quantitative by Migdal, hence the name Landau, Pomeranchuk, Migdal effect; LPM effect for brevity.

Until our experiment there was no quantitative test of the theory of the LPM effect. The theory used quantum mechanics and semi-classical electrodynamics to predict that the multiple scattering of the electron as it passed through dense matter would suppress the emission of low energy photons. Since no one today doubts the correctiveness of quantum mechanics or of semi-classical electrodynamics, why bother to carry out a quantitative test of the theory? Why did we do the experiment?

We did it because Spencer Klein thought of a neat way to do the experiment, we did it because we could use the SLAC linear accelerator in a parasitic way, not using expensive main beam time, we did it because it was a sweet experiment. As my paper reports, we have confirmed the prediction of Landau, Pomeranchuk, and Migdal, and incidentally we have pointed out that their formulation is imprecise and lacks generality. This in turn has stimulated new theoretical work by my colleagues, Richard Blankenbecler and Sidney Drell.

Still we found no new physics and it will be sometime before I work again on an experiment that is only neat and sweet. I enjoyed the experimental work on the LPM effect and I was pleased when I finally understood Migdal's tortuous mathematics in simple terms. I even had a vague hope that this experiment would stimulate me to think of ways to turn the experimental technique into something deeper. But back that hope melted away and the data analysis matured, my thoughts turned hard to my other deeper research work. Much of that research work is with the CLEO detector at the Cornell electron-positron collider CESR. With the large amount of data on tau lepton physics being collected with the CLEO detector, I am going to try again to find new physics through the tau lepton.

As I write this comment we are half way through the first experiment based on this paper and so I have the opportunity to write on an experiment in progress, something rarely done.

Previous general searches for free fractional electric charge have examined at most about 1 milligram of material per experiment. They found no evidence for free fractional electric charge, therefore the upper limit is less than 1 free fractional charge per 10²¹ nucleons. The usual candidate for a free fractional charge is an isolated quark with a charge ± e or ± e where e is the magnitude of the electron's charge.

The object of this paper is to describe how to increase the sensitivity of a fractional charge search by 10⁴ to 10⁵, searching through 10 to 100 grams of matter. But in our first experiment, in which we are developing the technology, we are examining about 1 milligram of matter. The experiment is being carried out mainly by Nancy Mar for her Stanford University PhD. research, by Eric Lee, and by me.

The method derives from the original work of Millikan and from free quark searches carried out at San Francisco State University. Small drops of liquid fall in air under the combined forces of gravity and a vertical electric field; the electric field changes sign periodically. At present we use oil drops with a 7 micrometer diameter which fall with an average velocity of 1.4 millimeters per second. The electric field changes sign every 0.1 seconds and the trajectory of the drops is observed over a vertical distance of about 2 millimeters.

The small diameter of the drops and the fractional resistance of the air leads to the drops falling with a constant velocity, usually called the terminal velocity. If the drop has non-zero electric charge, there are two different values of the terminal velocity, one value when the electric force adds to the gravitational force and the other value when the electric force subtracts from the gravitational force. Applying Stoke's law to these two velocities gives the electric charge on the drop and the mass of the drop.

The trajectory of the drops is measured using a CCD camera whose output is stored on a "frame grabber" board in a desktop computer. The computer calculates the charge and mass of each drop, stores the data, and operates the experiment…

I have little to add to this 1986 article which was in part a plea for greater professional and academic rewards for those who develop new techniques in accelerators and detectors for elementary particle physics research. In the past decade the rewards have increased, the notable example being the 1992 Nobel Prize in physics awarded to Georges Charpak for his particle detector discoveries.

But overall there is still too much attention given to fashionable theories in our journals and meetings and in the popular scientific press. There has been a shift in popularity to speculative particle astrophysics and cosmology, more about "dark matter" these days. Still some of the old favorites persist; without anymore evidence than we had in 1986, superstring theory still gets a headline every once in awhile.

With the final reprint in this volume I have indulged myself by combining my interest in physics with my interest in the history of mechanical and construction technology. I have a strong interest in the nineteenth and early twentieth centuries when mechanical invention and innovation reached its height. I collect mechanical antiques: farm machinery, old typewriters and sewing machines, acoustic phonographs. Finding old Meccano and Erector sets is a special pleasure. (I stay away from old toy electric trains, there's too big a crowd there.)

I had hoped to write more about the connections I see of physics with popular science and public technology, but I have not had the time, or not chosen to make the time. Perhaps I will make time in the future. For example, the best known and most popular toy construction sets used nuts and bolts to connect the parts: Erector in the United States, Meccano in France and the British Empire, Marklin in Germany. Yet toy inventors have known for a long time that nuts and bolts are hard to use, are easily lost, and discourage many children from playing with construction toys. There have been many attempts to devise boltless construction sets, some such as the Lionel construction set of the late 1940's have been manufactured. The only success has been LEGO, but LEGO copies a brick technology, not a steel girder technology. Here then is a subject that deserves a paper: mechanics and toy construction sets.

Please refer to full text.