In this paper a series of experiments is described relating to the deformation of the earth's crust by horizontal compressive stress. A floating model crust is compressed in most cases, after a slight dent has been run across it to represent a geosyncline. When materials of the correct order of strength are used, the crust buckles down below the model geosyncline forming a root at the lower surface of the crust. At no time in this process is a topographic depression formed at the surface that exceeds the depth of deep-sea troughs as compared to the thickness of the earth's crust. Other possibilities suggested by the experiments are that the crust may break through and overthrust and that a broad geosyncline will surmount a more complicated form of root. It was also found that the direction of compressive stress need not be at right angles to the course of the geosyncline to produce a root and that an orogenetic belt need not be more plastic than the remainder of the crust, but that the contents of the geosyncline must have less aggregate strength than a layer of the crust itself of the same thickness, as they would otherwise disappear down into the root. These experiments illustrate and clarify the theories of crustal buckling evolved by Vening Meinesz to account for the anomalies of gravity in the East Indies and by Bucher from geological data. In the second part of this paper an attempt is made to explain the recent isostatic history of the East Indies on the basis of the buckling hypothesis. The chief aim is to show that the anomalies may date back to the miocene orogenetic phase, and that the recent vertical movements can be looked upon as a slow adjustment to regain isostatic equilibrium. Geological Laboratory, Groningen.