All the non-biting flies examined, C. erythrocephala, M. domestica, S. carnaria, F. canicularis, L. caesar and O. anthrax, possess a filtering apparatus situated in the pseudo-tracheae of the suctorial disc. The anatomy and action of this filter have been most thoroughly studied in C. erythrocephala. The suctorial disc is grooved by pseudo-tracheae which end near its centre in closed collecting channels. The latter open into furrows or gutters formed by the peculiar disposition of the prestomal teeth on the walls of the prestomal cavity. The opening of the mouth is situated at the base of the cavity. During natural feeding the lobes of the suctorial disc are pressed together so that the lumen of the prestomal cavity is obliterated, and no food can enter the mouth except through the collecting tubes. The pseudo-tracheae are channels kept open by chitinous rings situated in their walls. Each ring has one bifid extremity, enclosing between the horns the “interbifid space”, which forms an opening of definite size into the pseudo-trachea. A fold in the cuticle, the “interbifid groove”, leads to each interbifid space.

The fluid food is sucked first along the interbifid grooves through the chitin-lined interbifid spaces into the pseudo-tracheae. Particles of larger diameter than the interbifid spaces (·006 mm.) are usually prevented from entering the mouth and are rejected. The fluid and smaller particles are drawn along the pseudo-tracheae, through the collecting channels and gutters between the prestomal teeth into the mouth. By means of strong suction two opposite interbifid grooves may be made to communicate with each other owing to the lateral fissures connecting with the longitudinal pseudo-tracheal fissure being forced open, and consequently a few larger particles, up to ·02 mm. in diameter, may be drawn into the pseudo-tracheae.

Certain relatively large and very attractive objects, such as the ova of tape-worms, too large to pass through the filter, may occasionally be swallowed. Such objects probably pass directly into the mouth, when the prestomal cavity is open, during the prolonged sucking efforts made by the flies.

The large number of experiments which have been made leave little room for doubt that under natural conditions, especially when the fly is feeding on a thin film of moisture, the filtering apparatus works with a high degree of efficiency.