For a wing with control devices in the form of inboard aft extensions of the planform, representing the tail of a bird, the aerodynamic characteristics have been analyzed. In the 1.3 m low-speed windtunnel of the Institut fur Stromungsmechanik at TU Braunschweig an experimental program has been carried out on an A = 5 rectangular Wing for a large variety of control surfaces attached to the trailing-edge. Length, width, area, lateral spreading as well as fork deepness of the tail have been altered systematically. In addition deflections-epsilon of the control surfaces up and down have been investigated, and for one planform shape also a twisted tail has been considered. Three- and six-component measurements have been carried out and the contributions of the various devices to stability and control have been determined. By adding small control surfaces to the wing longitudinal stability is improved. With increasing length, corresponding to increasing area, of the tail the longitudinal stability (- dc(m)/dc(L)) rises considerably whereas the control effectivenesses dc(L)/d-epsilon and dc(m)/d-epsilon remain constant. Lateral spreading of the tail without and with simultaneous increase of the control surface area leads to increased stability and control effectiveness. In forked tails the relative increase of stability and control effectiveness is larger than for unforked tails. In symmetrical flow the tail of a bird acts as a horizontal stabilizer and as an elevator in the same way as in conventional airplanes. Six-component measurements show for a twisted tail an increase Of lateral stability which is independent of the sense of rotation of the tail. On the other hand twisting of a loaded tail leads to a side force, a rolling and a yawing moment which depend on the sense of rotation in combination with the sign of the loading and which contribute to lateral control. In birds lateral stability and control is achieved in an unconventional manner by a twisted tail, which replaces the vertical fin and the rudder of conventional ai lanes.