The difference between a hypothetical and a theoretical approach is considered and the advantages in using a hypothesis is discussed. Virgin horizontal stresses play an important role in stabilising strata which bridge across an excavation, and reduces the tensile stresses which develop in a beam subjected to bending. Simple formulae are deduced for the bridging of a thick massive stratum under such conditions. At shallow depths, tensile failure on the upper side of the beam is the most likely cause of fracture, but below a certain depth this gives place to compressive failure on the underside, with the possible incidence of rock bumps. The presence of strong massive beds delay the onset of subsidence and may restrict the width of the subsidence trough. If the material involved in the caving is strong, this will reduce the final subsidence, especially at shallow depths. The British regulations for the control of aquifers and their dependence on the strain developed at the seabed are discussed. As an alternative, the concept of a disturbance factor is introduced to permit the consider- ation of aquifers at depth, leading to an explanation of the make of water on the first faces to be mined in the Selby Coalfield. 'Elie paper closes with a brief consideration of the different roles played by longwall face supports in the control of readily cavable strata, and the control of strong beds which hold up in the waste.