Two papers with quite different objectives established protocols that proved pivotal for future work on the role of gibberellins in seed germination. In their paper published in 1967, Russell Jones and Joseph Varner (Planta 72: 155-161) developed a bioassay based on induction of α-amylase activity in barley embryo-less half-seeds that was specific for bioactive gibberellins. The induction of α-amylase in the aleurone of barley and other cereals was to become the experimental system of choice to study gibberellin signalling. However, despite much progress in identifying the molecular events linking gibberellin action and α-amylase gene expression, in many cases their role in the process is still unclear. In 1987, Steven Groot and Cees Karssen (Planta 171:525-531) showed that germination of tomato seeds was limited by the ability of the radicle to penetrate the surrounding layers, with the endosperm forming the major barrier. They used a modified needle attached to a tensiometer to measure the force required to break through the endosperm. While in wild-type seeds, a factor from the embryo, assumed to be gibberellin, promoted breakdown of the endosperm, gibberellin-deficient seeds required an external supply of the hormone to weaken the endosperm or for it to be mechanically disrupted for germination to occur. The paradigm of seed germination being physically restricted by surrounding layers and the role of gibberellin in weakening these tissues has been confirmed in many eudicot species. Gibberellin signalling induces the production of cell-wall loosening enzymes in the micropylar endosperm adjacent to the radicle, but it is unclear whether or not this is a direct response. In both eudicot and monocot systems, there is still much to learn about the role of gibberellin signalling in germination.