Seafloor spreading anomalies in the world's ocean basins yieldcrucial information about the paleogeographic configuration of thecontinents and the development of ocean basins. Standard methodsfor measuring these variations are to apply ship towedmagnetometer systems or aircrafts. For the polar regions,magnetometer systems fixed to a vessel are more appropriate. Aprocessing sequence for the evaluation of magnetometer dataacquired onboard RV \emph{Polarstern} with a sensor system fixedto the ship is described in the first part of this thesis. In thesecond part, seafloor spreading anomaly data, measured by ship,aircraft and helicopter, in the South Atlantic region, are used toconstrain a geodynamic model for the Mesozoic break-up of Gondwanaand the development of the Weddell Sea Basin.Magnetometer data acquired onboard the German ice breaker andresearch vessel Polarstern have to be corrected for theship's interfering field caused by the hull and other steel partsof the ship. This correction is done by the calculation ofappropriate compensation coefficients. With the processing schemepresented in this thesis the ship's magnetometer data arecompensated and filtered to an accuracy of about 20nT at ahorizontal resolution of about 154m in quiet sea conditions.This makes the data suitable for a consistent interpretation ofsingle track data or further processing of a network of profiles.The compilation of a magnetic anomaly map and comparison withother high resolution magnetic anomaly data demonstrate thesuitability of the processed data to be used for detailedinterpretations of magnetic anomalies in geodynamic problems likethose encountered in the South Atlantic region and dealt with inthe second part of this thesis.A refined model for the break-up of Gondwana and the opening ofthe Weddell Sea is presented in the second part of this thesis.From an interpretation of a newly acquired extensive magneticanomaly data set in the eastern Weddell Sea, well constrained agesand orientations of seafloor spreading anomalies are used tocalculate a new set of model parameters for the opening of theWeddell Sea. These are combined with constraints on the SouthAmerica - Africa and Africa - Antarctica spreading systems, basedon a compilation of magnetic anomaly and fracture zone data, todetermine a new model for Gondwana break-up between 167Ma and83Ma. At the same time when north-south oriented separationbetween Africa and Antarctica is initiated by rifting in theSomali and Mozambique basins, stretching and extension takes placein the Filchner-Ronne Shelf and the Falkland Plateau between167Ma and 147Ma. In contrast to models proposing only rigidEast-West Gondwana separation, this model features three separatebasins developing simultaneously during early Gondwana break-up.These basins are the Somali, Mozambique, and Weddell Sea basins.In the new model, separation between South America and Antarcticastarts in the southern Weddell Sea at around 147Ma, about 15 to20 million years later than previously estimated. Subsequentspreading occurs at slow (~12-14mm/yr halfrate), and aftermagnetic chron M2 (122.3Ma) ultra-slow (~8mm/yr)spreading rates, with little change in the NNW spreadingdirection. Further consequences of this model are a revision ofthe age of formation of seaward dipping reflector sequences alongthe continental margin off western Dronning Maud Land and newconstraints on possible movements of the Mozambique Ridge betweenAfrica and Antarctica during the break-up of Gondwana and thedevelopment of a deep water connection between the Indian and theSouth Atlantic Ocean.