Static spherically symmetric solutions of a particular model of gravitation, including quadratic curvature and torsion terms, are discussed and compared with those found earlier in a model described by a different action. Considerable emphasis is placed on a Lorentz group covariant formulation of the models and the role of local diffeomorphisms of space-time carefully distinguished from the structure group for gravitation. It is argued that the 'modified Poincare gauge' approach of Hehl (1980) and co-workers mixes these concepts and that their model for gravity is dynamically specified by a locally Lorentz invariant action. The logical distinction between a space-time transformation generated by symmetries of particular solutions to a model and gauge covariant symmetries that apply to all solutions is demonstrated.