The OLYMPUS experiment was designed to measure the ratio between the positron-proton and electron-proton elastic scattering cross sections, with the goal of determining the contribution of two-photon exchange to the elastic cross section. Two-photon
exchange might resolve the discrepancy between measurements of the proton form factor ratio, $mu_p G^p_E/G^p_M$, made using polarization techniques and those made in unpolarized experiments. OLYMPUS operated on the DORIS storage ring at DESY, alternating between 2.01~GeV electron and positron beams incident on an internal hydrogen gas target. The experiment used a toroidal magnetic spectrometer instrumented with drift chambers and time-of-flight detectors to measure rates for elastic scattering over the polar angular range of approximately $25^circ$--$75^circ$. Symmetric M{o}ller/Bhabha calorimeters at $1.29^circ$ and telescopes of GEM and MWPC detectors at $12^circ$ served as luminosity monitors. A total luminosity of approximately 4.5~fb$^{-1}$ was collected over two running periods in 2012. This paper provides details on the accelerator, target, detectors, and operation of the experiment.
The roles played by mesons in the electromagnetic form factors of the nucleon are explored using as a basis a model containing vector mesons with coupling to the continuum together with the asymptotic $Q^2$ behavior of perturbative QCD. Specifically,
the vector dominance model (GKex) developed by Lomon is employed, as it is known to be very successful in representing the existing high-quality data published to date. An analysis is made of the experimental uncertainties present when the differences between the GKex model and the data are expanded in orthonormal basis functions. A main motivation for the present study is to provide insight into how the various ingredients in this model yield the measured behavior, including discussions of when dipole form factors are to be expected or not, of which mesons are the major contributors, for instance, at low-$Q^2$ or large distances, and of what effects are predicted from coupling to the continuum. Such insights are first discussed in momentum space, followed by an analysis of how different and potentially useful information emerges when both the experimental and theoretical electric form factors are Fourier transformed to coordinate space. While these Fourier transforms should not be interpreted as charge distributions, nevertheless the roles played by the various mesons, especially which are dominant at large or small distance scales, can be explored via such experiment--theory comparisons.
