Although taxonomists interpret the ovaries of most cacti as inferior,3the inferior position has been derived in a very different manner from that in most flowering plants. In the classical situation, the dorsal sides of connate carpels arc invested by a floral tube (hypanthium) ; in Cactaceac, on the other hand, the dorsal sides of the carpels are mostly free—not invested by a floral tube. The ventral sutures are pulled downward during formation of the hollow receptacle so that they come to lie on the sides of a locule between flanges of tissue which represent the reduced remains of fused margins and lateral walls of adjacent carpels. The morphology of an individual carpel is like that of a pea pod which has been split midway between dorsal and ventral bundles, then opened up so that the dorsal segment projects upward, the ventral segment containing the ovules, downward. In most cacti, the ventral segments are adnate to the walls of a hollow floral receptacle. In all cases observed so far, the floor of the locule appears not to be invested by carpellary tissue—it is interpreted here as representing free recoptacular tissue. In other words, the morphological bases of cactus carpels are not at the bottom of the locule, but above it, where the dorsal and ventral bundle systems diverge from the recurrent receptacular bundles. This interpretation, like the more complex one offered by Buxbaum, explains the commissural stigmas reported by Tiagi forOpuntiaandMammillaria.