Geochemical problems that would benefit from large scale experiments include: 1) the determination of the physical properties of the many common rocks that are chemically homogeneous on the scale of 5--100 cm but not on a scale of 1--5 cm, 2) the study of the influence of rock textures and compositions and of the chemistry of the pore solution on stress corrosion cracking, and 3) the study of chemical transport within gradients in temperature, chemical potentials and stress. The ultimate goal of a few well-chosen large scale studies should be the development of predictive models.