The behavior of the octopus when feeding upon shelled molluscs is described in four steps: selection of prey, boring a hole in the shell, secreting into the borehole, and pulling out the body oC the mollusc and eating it. Selection of some gastropod prey is determined by a firm, deep, either partial or complete obstruction in the aperture of the shell. Selection of other prey, gastropods, pelecypods, and amphineurans, may be based upon resistance by the mollusc to the application of force by the octopus. Drilling of the hole is done by the radula. Since the octopus will drill and secrete into empty shells with obstructions in the apertures, metabolites from a live mollusc are unnecessary. This technique can be usedto collect the secretion. The frequency of occurrence and the eflect of the secretion are discussed. The hypothesis that the octopus first tries by force to pull out the body and, failing to do so, drills a hole in the shell is experimentally supported.

The conditions under which the octopus drills or pulls out the body of the mollusc are incompletely understood. The octopus may drill two or more shells in succession without feeding upon the previously drilled and weakened snail. No two steps in the behavioral sequence are necessarily linked to each other. The drilling-feeding patterns are complex and plastic.