Haeckel's studies of development in calcareous sponges (1872) led him to develop the “Gastraea Theory,” which proposes that the ancestral mode of germ layer formation, or gastrulation, was by invagination to produce a functional gut. His observations that gastrulation in the Calcarea occurs by invagination of a ciliated larva upon settlement and metamorphosis were supported by remarkable photomicrographs of the stage by Hammer in 1908. Although no later work found the same stage, these concepts are repeated in texts today. We have re-examined embryogenesis and metamorphosis in Sycon sp. cf. S. raphanus in order to understand when gastrulation occurs. Almost all larvae settle on their ciliated anterior pole and metamorphose into a bilayered juvenile whose interior cells rapidly differentiate into choanocytes and other cells of the young sponge. After a four-year search we have found the transitory stage shown by Hammer in which the anterior cells invaginate into the posterior half of the larva. The hole closes and it is not until some days later that the sponge forms an osculum at its apical pole. To understand whether invagination comprises gastrulation and if the hole can be considered to be a blastopore we have carried out a review of the literature dealing with this brief moment in calcaronean sponge development. Despite the intrigue of this type of metamorphosis, we conclude that gastrulation occurs earlier, during formation of the two cellular regions of the larva, and that metamorphosis involves the reorganization of these already differentiated regions. Considering the pivotal position occupied by the Calcarea as the possible sister-group to all other Metazoa, these results call for a reassessment of germ layer formation and of the relationships of the primary germ layers among basal metazoan phyla.