In this thesis we study the properties of optical wire antennas. As experimental method for our investigations we use apertureless near-field optical microscopy. This technique achieves high spatial resolution well beyond the diffraction limit by utilizing the field enhancement at the apex of sharp tips. An interferometric measurement scheme allows us to detect both near-field intensity and optical phase. By using s-polarized light for illumination and detecting the p-polarized component of the backscattered light we are able to map the z-component of the electrical near- field. Optimizing polarizer and analyzer angles of our cross-polarization scheme ensures a background free plasmonic eigenmode mapping. By comparison with simulation data not including the tip we show that the measurement has little to no influence on the eigenmode. The samples investigated in this thesis are arrays of gold nano-wires prepared by electron beam lithography. We observe plasmon resonances in our near-field images as patterns of lobes and explain them by regarding the wires as one dimensional Fabry-Pérot resonators. The number of nodes in between the lobes is the resonance order. From eigenmodes well beyond quadrupolar order we extract both, propagation constant and reflection phase of the guided surface plasmon polariton with superb accuracy. The combined symmetry breaking effects of oblique illumination and retardation allow the excitation of dipole forbidden even-order resonances. By systematically varying the azimuthal illumination angle we are able to map the directional receiving and emission patterns of the wire antennas. In order to understand these patterns we develop an analytical model. In contrast to radio frequency (RF) antenna theories we not only assume surface currents but also take volume currents into account. The model also allows us to spotlight the differences between plasmonic and RF antennas. The equations we derive describe both, the property of the wires as resonators as well as the antenna emission / reception patterns. With just four – physically motivated – parameters we are able to fit measured as well as simulated data astonishingly well. With this model predicting the relative intensity and phase of the light absorbed and scattered by nano-wire antennas it has great potential for future research.