It has been well established that some of the features of the feeding and digestive process in molluscs are rhythmic in character. Notably Hirsch (1915, 1917, 1931) has demonstrated a rhythmic periodicity of secretion in the salivary glands and digestive glands of some carnivorous Gastropoda, and Krijgsman (1925, 1928) has done similar work on the land pulmonate Helix. In all of these animals—and in cephalopods, too, where there is a more elaborate nervous control of secretion—discontinuous feeding is the rule. In the other and perhaps larger category of molluscs—those continuously feeding on fine particles—the central mechanism of the gut is the crystalline style, or its forerunner the protostyle. Here the need is, in Yonge's words (1937), ‘to the extent to which they depend on extracellular enzymes for digestion, continuous secretion’. In Yonge's view, now classical, the style was regarded as ‘an ideal mechanism for the continuous liberation of small quantities of an amylolytic enzyme’. Graham (1939), in his work on style-bearing gastropods, showed that the style is in general confined to animals with a continuous feeding habit, whether by ciliary means or by using the radula to graze on and rasp off fine particles. It is known that style secretion stops and the style is frequentiy dissolved when animals are removed from the water and cease feeding (see Yonge, 1925).

SUMMARY. The primitive phonate family, the Ellobiidae, is discussed with reference to the structure, ecology and relationships of the following genera—Pythia, Ophicardelus, Cassidula, Ovatella, Leucophytia, Ellobium, Pedipes, Leuconopsis, Marinula, Rangitotoa, Melampus and Carychium. Ecologically, the family is shown to fall into four sections, a more primitive estuarine section, and sections containing marine intertidal, coastal terrestrial and inland terrestrial genera of rather more specialized character. In relation with detailed descriptions about to be published of the functional morphology of British ellobiids, a wider comparative account is now given of the digestive system and the genital ducts, and their evolution throughout the family. Further trends, chiefly concerning the adult and embryonic shell, are dealt with, in a discussion of the evolutionary pattern of the Ellobiidae. The frequent occurrence of parallel evolution, and the relative absence of adaptive radiation, or characters of obvious selective value, are noted. A scheme is proposed for the classification of ellobiid genera into subfamilies, which in some points departs from that of Thiele, based on radular studies by Odhner. A wider review is then given of the evolutionary significance of the Ellobiidae, with reference both to the evolution of pulmonates from prosobranchs, and to their relationships with primitive tectibranchs. Finally, the relationships of the Basommatophora and the Stylommatophora are discussed, and the significance of the Ellobiidae is pointed out, with reference to the origins of both groups. The Basommatophora are held to be a phylogenetically unnatural group. The Ellobiidae is the only family of this group that could have provided possible ancestors for Stylommatophora, and these are unlikely to have occurred among known genera. Reference is made to the geological evidence in support of the view of gastropod evolution presented from this study.