Of all anomopods, daphniids have been the most successful exponents of life in open water. Many of them are completely independent of the bottom and subsist entirely on seston. A few of them are truly planktonic. Although the family has been intensively studied from many points of view, various morphological attributes have remained either inadequately known or never investigated. Some of these attributes, understanding of which is necessary if functions are to be appreciated, are considered, especially in the genus Daphnia, with which other genera are later compared. They include aspects of general morphology, the exoskeleton, endoskeleton and muscular system. How Daphnia swims is described, antennal movements being analysed from high-speed cine films. Locomotion is clearly derived from a naupliar mechanism, though the nauplius has long been eliminated from the anomopod life cycle. Antennal beat is more versatile than is immediately apparent and the animals are capable of far more complex manoeuvres than the simple `hop and sink' movements in which they often indulge. The trunk limbs are responsible for collecting and manipulating the food. Their morphology and arrangement are discussed and their armature, especially as revealed by scanning electron microscopy, is considered. The armature of limbs 3 and 4 dominates the trunk limb complex and makes up an extensive filter chamber. The mouthparts and labrum are basically the same as those already described in detail for other anomopods, but the labrum lacks a keel. A wide range of particulate foods is consumed. A detailed account is given of the feeding mechanism, which has been studied both by direct observation and with the aid of high-speed cine-photography. Most of the basic principles involved were elucidated by Cannon, Storch and Eriksson who, however, disagreed on various points. The account now given is more detailed than any previously presented and is supported by numerous illustrations, whose lack has hitherto hindered comprehension. Parts of some of the earlier interpretations are incorrect, sometimes in ways that are not only intrinsically important, but which have led to erroneous views on such matters as the amount of energy expended in filtration. Trunk limb movements follow a regular rhythmic cycle. Water, containing suspended particles, flows into the carapace chamber via the ventral gape to replace that driven out posteriorly by the pumping action of trunk limbs 3 and 4 and their exopodites, is drawn into the filter chamber and through the filters borne on limbs 3 and 4 into interlimb spaces, from which it is finally expelled posteriorly. Trunk limb 5, whose movements initiate both promotion (the suction and filtration phase of the cycle) and remotion (the expulsion phase), seals the posterior interlimb space posteriorly during promotion of the limbs. There is no pressing of water through the filters during remotion of the limbs. Filtration occurs during approximately half the cycle. Notwithstanding claims to the contrary, the filter plates of trunk limbs 3 and 4 are correctly designated as such and serve as filters. Material abstracted by the filter plates is cleaned off by a series of devices, seven in all, passed into the median food groove, and swept forward by mechanical means to the mouthparts. The mandibles display a high degree of both skeletal and muscular asymmetry, which improves their performance. Any excess food material collected in the food groove is discarded. From the anterior end it is removed by the ejector hooks of the first trunk limbs, then swept out by the post-abdominal claws: from the posterior end it is removed by the post-abdominal claws alone. Errors and shortcomings in certain recent accounts that purport to explain the feeding mechanism are discussed. Trunk limbs 1 and 2 are incapable of filtration and are specialized for roles that have nothing to do with this process. The inapplicability of a model of filtration to the daphniid mechanism is noted and the importance of morphology, even in minute details, is emphasized. Contrary to recent suggestions, the function of `bristles' cannot easily be changed without changes in morphology. The necessity of understanding a mechanism before making calculations is emphasized and examples of misleading calculations, based on erroneous data, are noted. The habits of certain species of Daphnia are described. Both D. magna and D. obtusa are able to settle on their ventral carapace margins and attach themselves to surfaces, over which they can then glide forward, collecting food material by means of scraper-like spines borne distally on the second trunk limbs as they do so. D. magna can also lift accumulations of detritus from the bottom. Such material is then processed in the usual way. Some species sometimes indulge in swarming behaviour, which involves remarkable coordination between individuals. The way in which phenotypic changes in shape occur in Daphnia and the light this throws on phyletic changes in the genus are described, partly by the method of transformation of coordinates, which can be used to show changes in three dimensions, rather than the usual two. The influence of environmental factors is noted. Geographical, ecological and physiological aspects of radiation are considered. Other genera are treated more briefly. Daphniopsis departs little from Daphnia in its functional morphology and may not merit generic separation. Simocephalus attaches itself to a support by means of simple but effective specializations of the antennae and then remains stationary while it filters. This has enabled it to acquire a robust carapace in a way not permitted to Daphnia (of which a few of the more heavily built species sometimes rest on the bottom). Protection is thereby granted. Acquisition of this habit was probably assisted by the way in which Simocephalus swims, predominantly ventral surface uppermost. The feeding mechanism is essentially the same as that of Daphnia. Scapholoberis and Megafenestra have the same orientation during swimming as Simocephalus and have acquired the habit of hanging suspended beneath the surface film by their ventral carapace margins, for which they are highly specialized in morphology and behaviour. Here too the basic daphniid feeding mechanism is employed. Ceriodaphnia has specialized in small size. Although studied in less detail than Daphnia, it clearly has a similar feeding mechanism. Moina and Moinodaphnia are now often separated from the Daphniidae as the family Moinidae, but this seems unjustified. Trunk limb structure and the feeding mechanism are essentially the same as in other daphniids. These two genera, while primitive in certain respects, have a suite of specializations related to the nourishment of eggs and embryos by secretions produced by a Nahrboden, or `placenta'. This necessitates sealing of the brood pouch, by a device involving the post-abdomen, to prevent loss of the secretion. As embryos grow during development by the accretion of material from without, rather than from stored yolk, distortion and distension of the carapace are necessary to accommodate their increasing volume. The Daphniidae clearly arose from benthic ancestors, some indication of whose morphology and habits is given by certain extant macrothricids. Key features in the evolution of the family, which has existed since at least early Cretaceous times and probably originated even earlier than this, are listed. Of prime importance was the expansion of the gnathobasic filter plates of trunk limbs 3 and 4 at the expense of other filters.