Diverse field measurements and numerical modeling are used to address the problem of ice flow and sediment transport near the transition between cold-based and warm-based ice in a polythermal glacier terminus. With a two-dimensional finite element model, the impacts of different surface boundary conditions on ice flow trajectories are evaluated where ice flows over a slip/no-slip transition. Allowing ice to flow out of the top of the domain---explicitly forbidden in past theoretical treatments of the problem---significantly reduces stresses in the ice surrounding the transition. This modeling strategy was used with criteria for compressive ice fracture and frictional slip on existing fractures to determine conditions under which thrust faulting may occur in a glacier. The results indicate that compressive fracture of ice should be rare in glaciers, although slip is possible on existing fractures that are extensive, properly oriented, and weakened by elevated water pressure. Field measurements of englacial structure, surface and subsurface velocity, temperature, and stress were made at the terminus of Storglaciaren, a small polythermal glacier in northern Sweden. The formation of englacial debris bands in the northern part of the terminus has recently been ascribed to thrust faulting originating at an inferred slip/no-slip transition at the boundary between temperate and cold-based ice. Our field measurements indicate that the basal thermal transition (BTT) does not correspond to a slip/no-slip transition, although it may mark the start of a zone where water freezes to the glacier sole, thereby entraining sediment. Slip over a mostly weak bed beneath the accreted basal sediment limits longitudinal compressive stresses over most of the terminus. In the debris-laden northern part of the terminus, greater resistance to motion due to bed topography locally enhances longitudinal compression and causes upward transport of debris-laden basal ice. This upward transport is not, however, due to thrusting: strain rates are likely at least five orders of magnitude too small for fracture. Instead, structural and sedimentological observations along with stable isotope and tritium compositions in upglacier-dipping debris bands indicate that the debris-band sequence has been overturned by folding. The results highlight the role of studying ice mechanics and kinematics in a glacier terminus to constrain interpretations of glacial structural features and the landforms inherited from them.