EIGHT FIGURES
Recently, it has been shown that the vegetal plate is autonomous in respect to gastrulation (Moore and Burt, '39). Experiments carried out during the past summer have served further to localize the area of the gastral plate within which reside the forces producing the initial inpocketing, arid to show that the active region lies at the center of the plate. Measurements were also made which establish an approsimatc value for the force of invagination in the active area.
I n these experiments use has been made of the fact (Noore, '4.0) that the wall of the blastula, osmotically considered, is a mosaic consisting of two types of structure: (1) the cellular component which is relatively impermeable t o dissolved crystalloids, including salts ; (2) the intercellular substaiice lying between the cells which is freely permeable to salts but relatively irripermeable to organic molecules, especially disaccharides. As a consequence of this peculiar property of the intercellular component, it has been found possible to induce premature or pseudo-gastrulation in blastulae while they are still enclosed in their membranes, by the simple expedient of adding to the sea water a small amount of a solution of sucrose in a concentration isosmotic with sea water. The converse result is obtained if the larvae are reared in sucrose sea water, and later transferred to sea water. I n this case the sucrose molecules are caught in the blastocoele, and after the tenth segmentation, cannot escape. Hence, if, after this time, such