The anatomy of the tarsi and pretarsi of Phormia (Protophormia) terraenovae (R.-D.); Glossina palpalis (R.-D.) and Musca domestica L. is described with particular reference to structures possibly concerned in the action of deposits of contact insecticides. The numerous setae which cover the tarsal segments are distributed in a pattern which is common to all three species. Long spines situated at the distal extremity of each segment are the points of contact with which a fly engages a surface, and which take the thrust as the fly walks. Proximal to the primary spines are two ventral rows of chemoreceptors, protected from mechanical damage by smaller spines. In addition there are lateral and dorsal rows of bristles.

Contact chemoreceptors, which have been identified by experiment, are present in large numbers on the tarsi of P. terraenovae and M. domestica. Each chemoreceptor is a differentiated hollow seta possessing an extremely thin frontal membrane of cuticle, in which a lipoid layer appears to be incorporated and through which a lipoid-soluble insecticide might readily penetrate to the sensory neurocytes.

Receptors of this type are also present on the tarsi of G. palpalis, though tsetse flies have not been reported to possess a tarsal gustatory sense.

The pulvilli of all three species are entirely filled with viscous endocuticle and possess neither nerves, sense organs nor gland cells. The pulvilli are probably less important sites of penetration of insecticide from deposits than are the tarsel chemoreceptors, articulating membranes of setae, and intersegmental joint membranes.

Measurements of nerve diameters at different parts of the tarsi are used to interpret the results of Fisher (1952) concerning the action of DDT applied to limited areas of M. domestica. It is concluded that the relative toxicity of DDT at different parts of the body may be correlated with the number of sensory nerve fibres passing close to the site of penetration, but not with the number of sensory end organs directly affected.