No description available.

No description available.

MOND is an observational rule for predicting the acceleration of stars and galaxies from the distribution of the visible matter. It possibly stems from a new law of physics. I list the theoretical aspects of MOND, its achievements and problems. MOND has been tested mainly in disc galaxies so far. Its tests in elliptical galaxies are rare because the MOND effects are small for them in the parts observable by the conventional methods. In the thesis, I explain the methods and ideas I developed for testing MOND in the ellipticals using stellar shells. Moreover, the shells enable us to test MOND for stars in radial orbits for the first time. The shells are results of galactic interactions. I discuss the shell formation mechanisms and summarize the findings from shell observations and simulations.