We are conducting an imaging survey of pairs of elliptical galaxies which has already produced interesting results. Some pairs present a common pattern of distortion interpreted in terms of tidal effects (Davoust and Prugniel, 1988;. Other examples drawn from the literature (Borne and Hoessel, 1988;Colina and Perez-Fournon, 1990) share the same morphology. This Poster presents new cases and lists the characteristics of 24 such systems. Our pairs are drawn from a sample of binary and multiple galaxies which has in turn been extracted from the CGCG, UGC (Nilson, 1973) and VV (Vorontsov-Velyaminov, 1959) catalogues. This sample includes that of Karachentsev (1972). It contains 1800 pairs, among which 700 are S -S or mixed morphology pairs. We are working on the remainder to produce a sample of close physical pairs of elliptical galaxies (we also include bulge dominated SO's since the morphological discrimination from ellipticals is often ambiguous, in particular for interacting galaxies).One of the interests of this work is to provide a sample selected on purely optical criteria, at variance with other works (e.g. Valentijn and Casertano, 1988). This will allow statistical studies of non-optical properties of these pairs (in particular radio emission).We have so far obtained CCD-images of 125 pairs with the 2 m telescope of Pic du Midi in V and/or R. Velocities and velocity differences of 78 pairs have been obtained using the 1.93 meter telescope of Observatoire de Haute Provence and from the literature. One is an optical pair (VV 190). Eighteen of our pairs present the morphological effect described in Davoust and Prugniel (1988): the external parts of each member are stretched in opposite senses in a direction roughly perpendicular to the pair axis. The proportion of 15±4% distorted pairs confirms previous estimates.Except for a few cases involving flattened galaxies with nearly aligned major axes which deserve careful detailed analysis , the apparent distortions do correspond to physical distortions. We have searched the literature for isophote maps showing this effect. In the survey of radio galaxies by Colina and Perez-Fournon (1990), 7 out of 20 pairs show this characteristic distortion. Such a high rate (35%) suggests a relation between the strength of the interaction and radio activity (if we consider that the magnitude of the distortion is related to the strength of the interaction).On the other hand, among the optically selected distorted pairs, 42% are radio galaxies while only 5% of all E-E Karachensev pairs axe detected in radio (Stocke, 1978; this proportion rises to 7% if we only consider pairs with a separation of less than 10 kpc). This is not a rigorous statistical analysis since thresholds for radio detection and also for distortion detection are not 65