{include 4_1.doc|4.0 FUNDAMENTAL SYMMETRIES AND WEAK INTERACTIONS 4.1 CVC and SCC in the mass-8 isotriplet L. De Braeckeleer, E.G. Adelberger, K.A. Snover, D.W. Storm and D. Wright The conserved vector current (CVC) hypothesis predicts the existence of a small weak magnetism correction to the {EMBED EQUATION |}-decays of 8Li and 8B related to the isovector M1 and E2 decay strengths of the isospin analogue state in 8Be. By measuring the difference in {EMBED EQUATION |} angular correlations in the decays of these two nuclei,1 we may extract the quantity b/AC + dII/Ac, where b is determined by the M1 decay amplitude, c the GT decay amplitude, the first term is the CVC-predicted weak magnetism correction, and the second term represents second class current (SCC) contributions. In previous experiments in this laboratory,2 we have measured the GT strength of 8Li and 8B decay, as well as the isovector M1 strength of the decay of the isospin analogue doublet in 8Be, to excited 8Be. We have combined these measurements to produce a CVC prediction for the weak magnetism term, and compared this prediction to the {EMBED EQUATION |} correlation data of both Tribble and Garvey3 and McKeown, Garvey, and Gagliardi.4 The CVC prediction and the two data sets are shown in Fig. 4.1-1. If we assume CVC to be correct, we may allow for the possibility of an SCC contribution by fitting the difference between our prediction and the data. We obtain dII/Ac=0.0+0.3+0.3 from Tribble and Garvey's data, and dII/Ac=-0.5+0.2+0.3 from the data of McKeown, Garvey, and Gagliardi. The first error bar quoted represents the uncertainty in the {EMBED EQUATION |} correlation data, the second the uncertainty of our CVC prediction. If, on the other hand, we assume the absence of SCC, we may test the validity of CVC by comparing our prediction directly to the data. If we call the multiplicative difference between our prediction and the data {EMBED EQUATION |}, we obtain {EMBED EQUATION |}=1.00+0.04+0.05 for Tribble and Garvey's data, and {EMBED EQUATION |} = 0.93 +0.03+0.05 for the data for McKeown, Garvey, and Gagliardi. We thus conclude that present experimental data are consistent with both CVC and the absence of SCC.5 Uncertainties in the measured isovector gamma decay strength give a significant contribution to the uncertainty in the CVC prediction of the weak magnetism term. We are therefore planning improved measurements of both the total strength and its distribution to various excitation energies of 8Be. To this end, we have carried out a careful investigation of the sources of background in the 4He({EMBED EQUATION |}) experiment used to measure the gamma decay strength. Although difficulties in obtaining a correct normalization make it unattractive for a measurement of the total strength, the use of a long (~40cm) gas cell target which allows the center of the cell to be viewed by a detector while its Kapton windows are shielded by lead, eliminates an important source of {EMBED EQUATION |}-ray background and is therefore preferable for a measurement of the decay strength distribution. The remaining background may be attributed to {EMBED EQUATION |}-rays and neutrons produced when beam particles scattered by the windows or the target gas interact with the walls of the cell. By subtracting the {EMBED EQUATION |}-ray spectrum observed with the cell empty from the spectrum observed with a He-filled cell, we can compensate for the background produced by beam scattered from the cell windows, but not for background produced by beam scattered from the target gas. Crude estimates of the magnitudes of each of these backgrounds are consistent with experimentally observed spectra before and after subtraction. The remaining background after subtraction is unfortunately still comparable to the signal for {EMBED EQUATION |}-ray energies below ~8 MeV. To reduce this background, we have designed a new gas cell with a geometry chosen so that no {EMBED EQUATION |} particle singly scattered by the target gas hits the cell wall in a region visible to our detector with an energy of more than ~20 MeV. We plan to line these portions of the gas cell wall with Tantalum metal, whose Coulomb barrier will greatly inhibit background-producing interactions. The new cell will be used in a precision measurement of the gamma decay strength distribution. Fig. 4.1-1. The panels compare the +1{EMBED EQUATION |} CVC prediction (assuming no SCC), shown as a solid curve, to the data points of Tribble and Garvey (upper panel) and McKeown, Garvey, and Gagliardi (lower panel). } {include 4_2.doc|4.2 Measurement of the {EMBED EQUATION |} angular correlation in mass-8 E.G. Adelberger, J.F. Amsbaugh, P. Chan, L. De Braeckeleer, P.V. Magnus, D.M. Markoff, D.W. Storm, H.E. Swanson, K.B. Swartz, D. Wright and Z. Zhao Among the measurements needed to test CVC and/or search for second class current in the Mass-8 isotriplet, the measurements of the {EMBED EQUATION |} angular correlations in {EMBED EQUATION |} and {EMBED EQUATION |} are the most difficult to achieve with high accuracy. Both the weak magnetism and the weak electricity induced currents are proportional to the momentum transferred in the decay and therefore contribute to the observables at the percent level. The accuracy of these measurements determines the ultimate sensitivity of the symmetry tests. In particular, the {EMBED EQUATION |} energy dependence of the {EMBED EQUATION |} term in the {EMBED EQUATION |} angular correlation is presently quite uncertain. The two previous experiments1,2 support a large quadratic term, a fact not compatible with CVC and our recent measurement of the E2/M1 ratio, (see Section 4.1). Our apparatus has been designed with particular attention to the response function of the beta counters (delta E, active veto and stabilization). The data accumulated in August (1 week) and November 93 (3 weeks) have been analyzed. The {EMBED EQUATION |} angular correlation measured in {EMBED EQUATION |} is fitted by {EMBED EQUATION |} the {EMBED EQUATION |} and {EMBED EQUATION |} coefficients are shown in Figs. 4.2-1 and 4.2-2. The kinematical term {EMBED EQUATION |} shows a little deviation from the expected value at high energy. We are currently investigating the origin of this systematic effect. The {EMBED EQUATION |} coefficient does not have a significant quadratic energy dependence. Two additional {EMBED EQUATION |} counters have been built and set up at 0 and 180 degrees. This has required a modification of both the alpha counters and the hardware/software of the acquisition and stabilization system. We are now in the process of testing this new apparatus. We will also spend a couple of weeks adding to the existing statistics of the {EMBED EQUATION |} experiment. Finally, the {EMBED EQUATION |} run is awaiting the completion of the high intensity {EMBED EQUATION |} terminal ion source. Fig. 4.2-1. The kinematical coefficient {EMBED EQUATION |} measured Fig. 4.2-2. The {EMBED EQUATION |} coefficient measured in the {EMBED EQUATION |} in the {EMBED EQUATION |} decay. decay. } {include 4_3.doc|4.3 Completion of an apparatus to measure the PNC spin rotation of cold neutrons in a liquid helium target E.G. Adelberger, B.R. Heckel, D.M. Markoff, S.D. Penn and H.E. Swanson The apparatus to measure the parity non-conserving (PNC) spin-rotation of transversely polarized neutrons through a liquid helium target is nearly complete. This apparatus is designed to measure a neutron spin-rotation predicted to lie between 0 and 5 {EMBED EQUATION |} {EMBED EQUATION |} radians1,2 in 46 cm of helium with an error of 5 {EMBED EQUATION |} {EMBED EQUATION |} radians. This observable, in conjunction with {EMBED EQUATION |}({EMBED EQUATION |} + {EMBED EQUATION |}),3 will determine the PNC isovector pion-exchange amplitude of the NN interaction, {EMBED EQUATION |}. This amplitude is sensitive to the neutral weak current contribution. Current experimental results provide upper limits for {EMBED EQUATION |} that are smaller, by a factor of three, than the early theoretical calculations.4 Recent QCD sum-rule calculations5,6 predict a smaller value for {EMBED EQUATION |} that is consistent with existing data, and an expected rotation of 2 {EMBED EQUATION |} {EMBED EQUATION |} radians in our targets. Our experiment is motivated by the need for a more precise measurement of {EMBED EQUATION |}. The two coaxial {SYMBOL 109 \f "Symbol"}-metal shields have been constructed and tested. With the cryostat in place, we measured axial magnetic fields of 15-20 {EMBED EQUATION |}G in the center region, rising to approximately 40 {EMBED EQUATION |}G on the ends. This is comfortably below the maximum tolerable field of 100 {EMBED EQUATION |}G, set by the requirement that spin-rotations arising from diamagnetic effects of helium will be within our desired errors. The cryostat insert housing the liquid helium targets, the {SYMBOL 112 \f "Symbol"}-coil, and the pump and valve system that fill and empty the target chambers has been built. The feedthrough system which will drive the pump and valve will be completed shortly. The neutron detector, a segmented {EMBED EQUATION |} ionization chamber, is currently under construction. It is designed so that neutrons of different velocity ranges can be detected separately. The PNC effect is velocity independent, while spin-rotations from magnetic fields are velocity dependent. Velocity separated detection will allow us to monitor the integrated magnetic fields along the neutron path. In addition, the detector signal plates are divided into four quadrants that will allow us to monitor false signals from geometric asymmetries. The central {SYMBOL 112 \f "Symbol"}-coil has been constructed, wound, and assembled. Preliminary tests show the leakage fields to be on the order of {EMBED EQUATION |} near the coil, falling off to {EMBED EQUATION |} 5 cm away. The qualitative field shape was consistent with the sum of the coil symmetry (comparing favorably with computer predictions) and the winding asymmetry. The input and output coil forms have been constructed and are currently being wound with 1 mm wire. With {SYMBOL 109 \f "Symbol"}-metal pieces connecting the main coil with the return coils, we expect field homogeneities of a part in {EMBED EQUATION |}. The computer program to control the experiment and the data acquisition is currently being written and tested. Our scheduled beam time at the NIST (National Institute of Standards and Technology) reactor facility in Maryland will begin approximately two months after reactor start-up following the scheduled down-time for facility improvements. We expect to take data in the fall of 1995. 4.4 Measurement of Newton's constant G E.G. Adelberger, J.H. Gundlach, B.R. Heckel and H.E. Swanson Newton's constant G is one of the most fundamental yet least precisely known constants of nature {EMBED EQUATION |}1 and recently even this value has also been brought into question. The PTB (the German Bureau of Standards) obtained a value 0.6% (~40 standard deviations!) higher2 than the accepted value, a group from New Zealand reported a value which is 0.1% lower (~7 standard deviations)3 while another German group4 measured a value that is in accordance with the CODATA value. In addition a Russian group claims to observe a temporal variation of G at the 0.7% level.5 We are pursuing a novel and elegant technique that we believe may ultimately make a determination of G at the {EMBED EQUATION |}-level possible. We plan to use the new rotating torsion balance (described in Section 4.5) currently being developed for testing the universality of free fall. We will use the l,m = 2,2 gravitational coupling from two masses (30-70kg) on opposite sides of the pendulum to induce a torque on the torsion pendulum. The turntable rotation rate will be servoed so that the pendulum does not move with respect to the turntable i.e. so that the torsion fiber never twists. This essentially transfers the angular acceleration of the pendulum to the turntable. The angular acceleration will be measured using a high-quality shaft encoder attached to the turntable. We plan to use a flat vertical pendulum; the expression for the {EMBED EQUATION |}-moment of a two-dimensional pendulum contains the same integral over the mass distribution as does the moment of inertia. This largely avoids the important experimental problem of knowing the dimensions and density uniformities of the small pendulum. To eliminate spurious torques caused by masses in the laboratory we will rotate the attractor masses on a second turntable at a different and possibly opposite rotation rate. We made computer simulations to study this feedback scheme. From these we expect that a {EMBED EQUATION |} acceleration measurement would ultimately be possible. We already implemented the rotation feedback in our existing rotating balance and were able to verify the computer simulations. The dominant noise source we encountered was due to gravitational effects from pedestrian and vehicular traffic in the vicinity of the apparatus. The site of the new rotating balance is in a relatively isolated spot on the campus and should reduce this problem. Furthermore the rotation rate of the attractor masses can be chosen to give a fairly high signal frequency which will reduce gravitational 1/f-noise. 4.5 Construction of a new rotating torsion balance instrument E.G. Adelberger, J.H. Gundlach, B.R. Heckel, S. Penn, Y. Su and H.E. Swanson We are preparing a new, more sensitive rotating torsion balance apparatus to search for violations of the Equivalence Principle over length scales ranging from 1 m to infinity. The dominant limitations of our previous results arose from: {SYMBOL 183 \f "Symbol"} Brownian motion of the torsion pendulum due to gas damping in the residual vacuum of {SYMBOL 187 \f "Symbol"}0.1 Torr. {SYMBOL 183 \f "Symbol"} coherent imperfections in the turntable such as fluctuations in the rotation rate or vertical "rumble''. {SYMBOL 183 \f "Symbol"} residual gravity-gradient couplings. {SYMBOL 183 \f "Symbol"} daily variations in the tilt of the laboratory floor that required corrections to the data. {SYMBOL 183 \f "Symbol"} vertical seismic motion was suspected (but never proved) to contribute to our fluctuating errors. Our new instrument is designed to minimize the first four of these problems. {SYMBOL 183 \f "Symbol"} an ion-pump will be used to evacuate the chamber to {EMBED EQUATION |}torr. {SYMBOL 183 \f "Symbol"} the torsion balance will be rotated on a high-quality air-bearing turntable on which a state- of-the-art angle encoder and an eddy current motor are mounted directly. The turntable system is being manufactured by a commercial firm, and is scheduled for delivery in late spring. {SYMBOL 183 \f "Symbol"} the tilt-sensitive parts of the apparatus will hang from a 2-axis gimbal constructed using flexures to avoid sticking and hysterisis. {SYMBOL 183 \f "Symbol"} the pendulum will be more symmetric; it will have 8 testbodies that mate reproduceably in conical seats. {SYMBOL 183 \f "Symbol"} the gravity gradient compensation will be improved by placing the compensators farther from the balance. We will develop the apparatus in two phases. First we will upgrade our existing Eöt-Wash balance for high vacuum operation, and hang it from a gimbal mechanism attached to the new turntable. This will allow us to thoroughly test the turntable and identify the necessary improvements for the second phase. We have built a massive concrete platform 5 m above the floor of the old cyclotron cave on which the turntable will be mounted and below which the apparatus will hang. The pendulum will be 2.5m above the floor to reduce the ambient {EMBED EQUATION |} gravitational gradient. This initial setup should yield a factor of 5 more precise test of the Equivalence Principle. The second phase will have a gimbaled structure inside the vacuum vessel to support the tilt-sensitive components and a much longer torsion fiber. We will explore using a liquid-nitrogen cooled jacket to reduce thermal noise and fiber relaxation noise. With this second phase we hope to probe the Equivalence Principle with a substantially improved sensitivity. 4.6 Progress with the rotating-source torsion balance experiment E.G. Adelberger, J.H. Gundlach, M.G. Harris, B.R. Heckel, G.L. Smith and H.E. Swanson In last year's Annual Report,1 we reported preliminary results that were apparently limited by systematic errors from gravity gradients arising from imperfections in the 3 ton Uranium source. The largest systematic error arose from the l, m=3,1 coupling. {SYMBOL 183 \f "Symbol"} the stray {EMBED EQUATION |} gradient of the source (due to machining imperfections) could not be measured precisely (and therefore minimized) because the {EMBED EQUATION |} gradient of the earlier source did not vanish by design. {SYMBOL 183 \f "Symbol"} the coupling to the stray {EMBED EQUATION |} moment of the pendulum (due to small test body misplacements) led to variations larger than the statistical error. We therefore modified our 3 ton Uranium source mass to have vanishing {EMBED EQUATION |}and {EMBED EQUATION |} moments. This was accomplished by introducing two horizontal gaps above and below the midplane of the Uranium brick stacks. We used flat aluminum bearing plates as spacers; these allow us to rotate the central stack by 180{SYMBOL 176 \f "Symbol"} into a configuration which maximizes the {EMBED EQUATION |} moment. Using this source configuration to measure the stray {EMBED EQUATION |} moment of the pendulum and special {EMBED EQUATION |} test bodies to measure the stray {EMBED EQUATION |} gradient of the source in its normal configuration, we were able to demonstrate that most of the offset (torques that are independent of the test body configuration on the pendulum tray) is due to the {EMBED EQUATION |}-coupling. We have tuned the source and the pendulum so that {EMBED EQUATION |} offset torques are small, and we verified that the corrections due to {EMBED EQUATION |} coupling are negligible. In addition we have built a set of new Pb and Cu precision test bodies that seat more reproduceably in the pendulum tray. We now operate our instrument at high vacuum and take data with the source mass rotating faster ({EMBED EQUATION |} than before and we achieve a statistical error of 5nrad{EMBED EQUATION |}. We have made several tests for non-gravitational systematic sources of error, and established that magnetism, thermal variations, and apparatus tilt now lead to insignificant uncertainties. To reduce our thermal sensitivity we installed a glass tube which surrounds the fiber. The Ag coated tube increases the pendulum time response to temperature changes to approx. 10h. } {include 4_7.doc|4.7 New tests of the universality of free fall E.G. Adelberger, M. G. Harris, B. R. Heckel, G. Smith and Y. Su Our tests of the Universality of Free Fall (UFF) have reached the practical limits of the current version of the Eöt-Wash torsion balance. We studied differential accelerations of Be-Cu and Be-Al test-body parts in the fields of Earth, the Sun, or Galaxy, and in the direction of the cosmic microwave dipole. We also compared the acceleration towards the Sun and our galactic center of Cu and single-crystal Si in an Al shell (this pair of bodies approximates the elemental compositions of Earth's core and the Moon or Earth's crust, respectively). In terms of the classic UFF parameter {EMBED EQUATION |}, our Earth-source results are {EMBED EQUATION |} and {EMBED EQUATION |} where all errors are 1{EMBED EQUATION |}. Thus our limit on UFF violation for Be and a composite Al/Cu body is {EMBED EQUATION |}. Our solar-source results are {EMBED EQUATION |}, {EMBED EQUATION |}, and {EMBED EQUATION |}. This latter result, when added to the lunar laser-ranging results that senses both composition-dependent forces and gravitational binding-energy anomalies, yields a nearly model-independent test of the UFF for gravitational binding energy at the 1% level. A fivefold tighter limit follows if composition-dependent interactions are restricted to vector forces. Our galactic-source results test the UFF for ordinary matter attracted toward dark matter, yielding {EMBED EQUATION |}, {EMBED EQUATION |}, and {EMBED EQUATION |}. This provides laboratory confirmation of the usual assumption that gravity is the dominant long-range interaction between dark and luminous matter. We also tested Weber's claim that solar neutrinos scatter coherently from single crystals with cross sections {EMBED EQUATION |} times larger than the generally accepted value and rule out the existence of such cross sections. These results have recently appeared in print.1 We are now designing a new instrument with improved sensitivity (discussed elsewhere in this report). } {include 4_8.doc|4.8 Search for {EMBED EQUATION |} rays following the {EMBED EQUATION |} decay of {EMBED EQUATION |} to the first excited {EMBED EQUATION |} state of {EMBED EQUATION |} L. De Braeckeleer, M. Felton* and A. Poon A natural attempt to reduce the background of a very low counting rate measurement is to set up a coincidence experiment. It has been known for a long time that the ultimate sensitivity of a {EMBED EQUATION |} {EMBED EQUATION |} {EMBED EQUATION |} decay search could exceed the one of {EMBED EQUATION |} because the photon deexcitation provides an additional signature that can be used to reject the background. Moreover, the physics of a {EMBED EQUATION |} {EMBED EQUATION |} {EMBED EQUATION |} decay has its own particular features. The neutrinoless {EMBED EQUATION |} transition is extremely interesting for particle physics since its observation would imply both a finite value of the neutrino mass and the existence of a right handed current. It is also interesting for nuclear physics because it involves the {EMBED EQUATION |} nucleon transition, a process strictly forbidden for the {EMBED EQUATION |} case. However, the pessimistic estimates of the rates for {EMBED EQUATION |} transitions due to both phase space suppression and small nuclear matrix elements have limited the enthusiasm of experimentalists for the search of this process. Is it possible to use the experimental advantage of the extra signature of the {EMBED EQUATION |} of the nuclear deexcitation and to avoid the theoretical disadvantage of the small matrix elements governing the rate of {EMBED EQUATION |} transitions? In the long wavelength approximation, the double beta decay operators can connect an initial {EMBED EQUATION |} state with {EMBED EQUATION |}, {EMBED EQUATION |}, {EMBED EQUATION |}states in the daughter nucleus. As a general rule, a decay to the {EMBED EQUATION |} is kinematically forbidden due to its higher excitation energies. In some cases, the decay to the first excited {EMBED EQUATION |} state is allowed. A favorable case is {EMBED EQUATION |}, with a Q value of 2 MeV to the first excited {EMBED EQUATION |} state in {EMBED EQUATION |}. Recently, two groups have attempted the observation of the {EMBED EQUATION |} decay of {EMBED EQUATION |} to the first excited {EMBED EQUATION |} state of {EMBED EQUATION |}. Assuming similar matrix elements as the ones governing the transition to the ground state, one expects a partial half-life of {EMBED EQUATION |} years. Presently the two groups report conflicting results: {EMBED EQUATION |} years1 and a null result at the level of 2 {EMBED EQUATION |} {EMBED EQUATION |} years.2 Previous experiments have focused on a very low background, special materials and underground laboratory. We are investigating a different approach: the detection in coincidence, of the 2 {EMBED EQUATION |} rays following the {EMBED EQUATION |} decay of {EMBED EQUATION |} to the first excited {EMBED EQUATION |} state of {EMBED EQUATION |}. As a preliminary test, we are using 2 medium size (55%) BGO suppressed Germanium detectors (side mounted). We are measuring the coincidence background between these 2 counters. The BGO's are used to veto the cosmic ray background as well as the natural radioactivity. The apparatus is covered by a 4" thick layer of OFHC copper and a 4" thick layer of lead. The background level at the energies of 540 and 590 keV is (0.02 {EMBED EQUATION |} 0.01) count per (1 keV)2 per year. We are now measuring the coincidence background with a small sample of molybdenum (35 g.) to find out how the apparatus responds to the radioactivity inserted in the sample itself. A Monte Carlo calculation of the efficiency of an apparatus equipped with 2 large detectors (170%) is under development. } {include 4_9.doc|4.9 Test of time reversal symmetry: The emiT experiment S.R. Elliott, R.G.H. Robertson, T.D. Steiger, D.I. Will and J.F. Wilkerson The fact that CP (combined charge conjugation and parity symmetry) conservation is violated in kaon decays was discovered over three decades ago.1 Despite the considerable efforts of numerous researchers, however, this phenomenon remains to be adequately explained. The observation of CP symmetry violation combined with the CPT theorem implies that T (time reversal) symmetry must also be violated. Thus, tests of T symmetry provide a probe of the thirty-year-old CP puzzle which complements direct CP symmetry tests. Scientists at the Nuclear Physics Laboratory along with colleagues from Los Alamos National Laboratory, the National Institute of Standards and Technology (NIST), Notre Dame University, the University of California at Berkeley/Lawrence Berkeley National Laboratory, and the University of Michigan, have formed the emiT Collaboration to carry out the most precise test of T symmetry ever performed using neutron decay. The complete neutron beta-decay distribution may be written: {EMBED EQUATION |} where {EMBED EQUATION |} and {EMBED EQUATION |} are the momentum and energy of the electron, {EMBED EQUATION |} and {EMBED EQUATION |} represent the emitted neutrino, {EMBED EQUATION |} is the spin direction of the neutron, and {EMBED EQUATION |} includes phase-space factors and the Fermi function. In this equation, the term proportional to the triple correlation {EMBED EQUATION |} is the only term which is odd under time reversal. Thus, the goal of the emiT experiment is to measure or place limits on the coefficient, D, which describes the strength of time reversal violation. The experiment will be performed by observing in-flight decay of low-energy (<10 meV) neutrons from the Cold Neutron Research Facility at NIST in Gaithersburg, MD. Using energy and momentum conservation, the unobservable variables describing the neutrino may be replaced by measurable variables describing the proton. Thus, D may be written in terms of {EMBED EQUATION |}, and this quantity may be measured by detecting both the electron and the proton, and monitoring the angular correlation between their momenta as the neutron polarization is flipped. The emiT detector consists of four plastic scintillator paddles for electron detection and four arrays of large-area PIN diodes to detect the protons. These eight detector segments are arranged in an alternating octagonal array about the neutron beam so that each segment of one type lies at an angle of 135ş relative to two segments of the other type. This geometry takes advantage of the fact that the electron-proton angular distribution is strongly peaked due to the disparate masses of the decay products. The emiT experiment is currently concluding the design phase and entering the construction phase. Assembly on the floor of the NIST reactor will begin before the end of 1995. The primary responsibility of the UW team is the production of the proton detector segments including read-out electronics, detector support frame, and associated vacuum systems. The feasibility of the proposed proton detection scheme was decisively proven during a test run at the NIST reactor in 1992.2 The proton detector segments and the front-end electronics are currently under construction at the Nuclear Physics Laboratory. } 1Nuclear Physics Laboratory Annual Report, University of Washington (1994) p. 19. 2Nuclear Physics Laboratory Annual Report, University of Washington (1994) pp. 20-21. 3R.E. Tribble and G.T. Garvey, Phys. Rev. C 12, 967 (1975). 4R.D. McKeown, G.T. Garvey and C.A. Gagliardi, Phys. Rev. C 22, 738 (1980). 5L. De Braeckeleer et al., Phys Rev C, in press. 1Tribble et al., Phys. Rev. C 12, 967 (1975). 2McKeown et al., Phys. Rev. C 22, 738 (1980). 1Y. Avishai, Phys. Lett. 112B, 311 (1982). 2V.F. Dmitriev et al., Phys. Lett. 125, 1 (1983). 3J. Lang et al., Phys. Rev. C 34, 1545 (1986). 4E.G. Adelberger and W.C. Haxton, Ann. Rev. Nucl. Part. Sci. 35, 501 (1985). 5E.M. Henley, private communications. 6G. Feldman et al., Phys. Rev. C 43, 863 (1991). 1CODATA (Committee on Data for Science and Technology) 1986, based on "1986 Adjustments of the Fundamental Physical Constants" By E.R. Cohen and B.N. Taylor, Rev. Mod. Phys. 50, 1121 (1987). 2W. Michaelis et al., presented at the Conference on Precision Electromagnetic Measurement, Boulder, CO, 27 June - 1 July, 1994. 3M. Fitzgerald and T. Armstrong, presented at the Conference on Precision Electromagnetic Measurement, Boulder, CO, 27 June - 1 July, 1994. 4H. Meyer, presented at the Seventh Marcel Grossmann Meeting on General Relativity, Stanford, July, 1994. 5V.P. Izmailov et al., Measurement Techniques 36, no. 10 (1993). 1Nuclear Physics Laboratory Annual Report, University of Washington (1994) p. 28. 1Y. Su et al., Phys. Rev. D 50, 3614 (1994). *Department of Physics, University of Washington, Seattle, WA 98195. 1A.S. Barabash et al., Nucl. Phys. B, S28A, 236 (1992). 2D. Blum et al., Phys. Lett. B 274, 506 (1992). 1J.H. Christenson et al., Phys. Rev. Lett. 13, 138 (1964). 2E.G. Wasserman, Ph.D. thesis, Time Reversal Invariance in Polarized Neutron Decay, Harvard University, 1994. {PAGE|34}