This paper describes a number of displays of various gulls (Laridae) with special reference to the Herring Gull group, the hooded gulls, and the Kittiwake, and discusses their functions, causation, evolutionary origin and further evolution as signals. After a sketch of the rationale of comparative behaviour studies (chapter II), the most common single displays and display sequences are described (chapter III). They are rather similar through the family, though many species-specific differences exist. The behaviour similarities strengthen the conclusion that the gulls are a monophyletic group. Evidence is presented in chapter IV to show that the displays have signal function. A distinction is made between distance-increasing and distance-reducing displays. The reasons for the occurence of more than one, yet no more than about six distance- increasing postures are discussed ; it is argued that each display may well be adapted to deal with a distinct category of opponent : accidental trespassers require another repellent than intentional persistent intruders; and within each category actual trespassers are met in another way than potential trespassers. In addition, one particularly loud call is a typical long-distance threat. This call usually acts as advertisement in a double sense: it repels competitors and attracts unmated females. Some of the distance-reducing (or appeasement) postures are used in agonistic situations, and still more regularly at the end of the pair formation or meeting ceremony. The need for and the possibility of more systematic studies of the precise functions of displays is stressed. Chapter V discusses aims and methods of analysis of the motivation of the displays. The value of "natural experiments" is stressed. The application of three independent methods leads to the conclusion that agonistic displays are ambivalent, i.e. the outcome of the simultaneous arousal of a tendency to attack and a tendency to flee: the relation between fluctuations in these tendencies and the displays shown is examined. There is a striking correlation between the motivation of these displays and the information they pass on to other individuals. Appeasement gestures always contain an element of fear; this tendency is in conflict with a tendency to stay, which can be, but usually is not part of the tendency to attack; it may be sexual attraction, attraction to at nest site, or attraction to a provider of food. The need for motivational analyses of the many different forms in which one posture can occur is stressed. The similarity of motivation in the pair formation ceremonies in the diff.erent species is much greater than the formal similarity of the display sequences; the conflict theory of courtship is tested. The origin of the displays (chapter VI) is varied. Some have clearly arisen as preparatory or intention movements of the patterns directly aroused by the situation ("aut- ochthonous" movements) ; of these, some are redirected to inanimate objects. Others are derived from movements belonging to functional patterns not directly aroused by the situation ("displacement activities") ; their various origins are discussed, and it is shown that they are second components of a dual movement, of which the first component is "auto?hthonous" in the above sense and facilitates the displacement activity. In chapter VII some ultimate causes of evolutionary change are discussed, and a preliminary functional classification of alleged changes in displays is presented. It is argued that change has been enhanced by at least four different types of selection pressure: (i) towards improvement of signal function (conspicuousness) ; (2) towards increase in intraspecific unambiguity; (3) towards increased interspecific unambiguity; and (4) as a corrollary of selection pressure in other functional systems. In some gulls, (4) may have made the major contribution to sexual isolation. A preliminary classification is given of the postulated evolutionary changes in behaviour mechanisms. The value of behaviour characters for taxonomic use is considered. After a discussion of adaptive and non-adaptive differences the relative validity of LORLNz's emphasis on the phylogenetic conservatism of displays is reconsidered, and the occurrence of convergent similarities is demonstrated.