The posterior attachment organ (sucker) of Temnocephala sp. is located ventrally attached to the posterior end of the body by a well defined stalk; those of Udonella caligorum Johnston and Anoplodiscus cirrusspiralis Roubal, Armitage et Rohde are extensions of the posterior end facing posteriorly. In Philophthalmus, the sucker is ventrally embedded in the main body. The sucker of Temnocephala is lined by an epidermis, its ventral part separated from the adjacent epidermis by a septate junction. The epidermis resembles that of the body proper, containing nuclei and numerous dense bodies, its surface enlarged by short microvilli, traversed by glandular ducts of two types and by sensory receptors, and based on a basal lamina with a thick underlying fibrous matrix. The stalk of the sucker contains many muscle fibres extending from the main body into the sucker. The posterior surface of the sucker of Udonella is separated from the adjacent tegument by a septate junction; it consists of numerous microvilli arising from the basal lamina and does not represent a tegument; glandular ducts of two types open through it, and muscle fibres extend from the body proper into the sucker. The posterior surface of the sucker of Anoplodiscus consists of a thin tegument not separated from the adjacent tegument by a septate junction, drawn out into a very large number of densely packed, long microvilli, some branching from a thick cross-striated base; large glandular ducts open postero-laterally. The ventral sucker of Philophthalmus is embedded in the body proper but clearly bounded by a “capsule” of basal lamina; it is lined by a tegument continuous with that of the main body and lacking microvilli except in a small band around the ventral sucker opening. There is no evidence from ultrastructure that the stickers of the four taxa are homologous. Since there is no convincing other evidence for the homology of the posterior attachment organs of the major groups of parasitic Platyhelminthes (Neodermata) and the Temnocephalida, a “cercomer theory” assuming such homology cannot be accepted as proven.