We have extracted QCD matrix elements from our data on double polarized inelastic scattering of electrons on nuclei. We find the higher twist matrix element \tilde{d_2}, which arises strictly from quark- gluon interactions, to be unambiguously non zero. The data also reveal an isospin dependence of higher twist effects if we assume that the Burkhardt-Cottingham Sum rule is valid. The fundamental Bjorken sum rule obtained from the a0 matrix element is satisfied at our low momentum transfer.

We have examined the spin structure of the proton in the region of the nucleon resonances (1.085 GeV < W < 1.910 GeV) at an average four momentum transfer of Q^2 = 1.3 GeV^2. Using the Jefferson Lab polarized electron beam, a spectrometer, and a polarized solid target, we measured the asymmetries A_parallel and A_perp to high precision, and extracted the asymmetries A_1 and A_2, and the spin structure functions g_1 and g_2. We found a notably non-zero A_perp, significant contributions from higher-twist effects, and only weak support for polarized quark--hadron duality.

The photoproduction of pi0pi0 and pi0pi+/- pairs off 40Ca has been investigated with the TAPS detector using the Glasgow photon tagging spectrometer at the Mainz MAMI accelerator. Data have been taken for incident photon energies in the energy range from 200 - 820 MeV. Total cross sections have been extracted from threshold up to the maximum photon energy and invariant mass distributions of the pion pairs have been obtained for incident photon energies between 400 - 500 MeV and 500 - 550 MeV. The double pi0 invariant mass distributions show some relative enhancement with respect to the mixed charge channel at small invariant masses. The effects are smaller than previously observed for lead nuclei and the distributions do not significantly deviate from carbon data. The data are in good agreement with the results of recent calculations in the framework of the BUU model, with careful treatment of final state interaction effects but without an explicit in-medium modification of scalar, iso-scalar pion pairs. This means that for Ca most of the experimentally observed effect can be explained by final state interactions. Only at low incident photon energies there is still a small low mass enhancement of the double pi0 data over the BUU results.

Recent experimental results for meson photoproduction from nuclei obtained with TAPS at MAMI are analyzed in view of the suppression of the second nucleon resonance region in total photoabsorption. The cross sections can be split into a component from the low density surface region of nuclei and a component which scales more like the nuclear volume. The energy dependence of the surface component is similar to the deuteron cross section, it shows a clear signal for the second resonance peak assigned to the excitation of the P11(1440), D13(1520), and S11(1535). The volume component behaves differently, it is lacking the second resonance peak and shows an enhancement at intermediate photon energies.

Photoproduction of neutral pions from nuclei (carbon, calcium, niobium, lead) has been studied for incident photon energies from 200 MeV to 800 MeV with the TAPS detector using the Glasgow photon tagging spectrometer at the Mainz MAMI accelerator. Data were obtained for the inclusive photoproduction of neutral pions and the partial channels of quasifree single pi0, double pi0, and pi0pi+/- photoproduction. They have been analyzed in terms of the in-medium behavior of nucleon resonances and the pion - nucleus interaction. They are compared to earlier measurements from the deuteron and to the predictions of a Boltzmann-Uehling-Uhlenbeck (BUU) transport model for photon induced pion production from nuclei.

Differential cross sections of the reactions $(\gamma,\pi^\circ\pi^\circ)$ and $(\gamma,\pi^\circ\pi^++\pi^\circ\pi^-)$ have been measured for several nuclei ($^1$H,$^{12}$C, and $^{\rm nat}$Pb) at an incident-photon energy of $E_{\gamma}$=400-460 MeV at the tagged-photon facility at MAMI-B using the TAPS spectrometer. A significant nuclear-mass dependence of the $\pi\pi$ invariant-mass distribution is found in the $\pi^\circ\pi^\circ$ channel. This dependence is not observed in the $\pi^\circ\pi^{+/-}$ channel and is consistent with an in-medium modification of the $\pi\pi$ interaction in the $I$=$J$=0 channel. The data are compared to $\pi$-induced measurements and to calculations within a chiral-unitary approach.