Since 1900, several scholars have argued that the terms "Olympian" and "chthonian" are commonly misused or overused, and that in the realm of ritual in particular the difference between sacrifices with and those without participation in the offerings by the worshipers does not coincide with the difference between Olympian and chthonian divinities. Fritz Graf and Walter Burkert, applying a model from social anthropology, have lately maintained that participation and nonparticipation are "ritual symbols," that is, variables employed among others to articulate (...) phases within the ritual itself; they imply nothing about any recipient, and have to do only with "the inner logic of the ritual." The present paper undertakes a reassessment of the relationship between recipients of sacrifice and the various sacrificial modes from the point of view of the Olympian/chthonian distinction. It argues that Olympian and chthonian sacrificial modes are clearly distinguishable, and that the character of the divine recipient is a fundamental constitutive element of Greek ritual. The basic principles of the author's approach are worked out with reference to the test case of rituals attested at various places and dates for Zeus Polieus. It is suggested that there is a remarkable consistency of specific ritual motifs in all these cases; that very specific conceptual themes and areas of interest, centering on agriculture, are everywhere associated with this god; and that the rituals and the themes cohere with one another and constitute a specific application of the Olympian/chthonian distinction predicated on the special characteristics of this particular divinity, who has a foot in each realm. It is argued that the Olympian/chthonian distinction retains its basic significance if it is applied in a less mechanical way than it has traditionally been. It is a central organizing principle in Greek religion, but does not represent a sufficient basis for analyzing individual divinities or rituals: specific character traits and interests and the circumstances of particular rites are fundamental, and will affect its application in given cases. Nor are the two categories mutually exclusive: they constitute one essential system at work in shaping the phenomena of Greek religion, but there is a much larger area of intersection between the two sets than has generally been recognized. One specific element of Greek practice-sacrifices with participation, but where the participation is required to take place in the sanctuary-are studied in detail, and it is suggested that they belong to an area of ritual intersection between the Olympian and chthonian categories. Recipients of such sacrifices are wholly or partly chthonian in character ; the desire not to destroy meat has led in these cases to a variation on holocaust sacrifice in the direction of Olympian banqueting: participation, but in a tightly controlled ritual setting. A Hebrew parallel for this sort of ritual compromise is suggested. On the basis both of the study of this particular sacrificial mode and of the more fluid approach to the general distinction sketched earlier, a reconsideration of some canonical lists of rites regarded as exceptions to the Olympian/chthonian distinction is undertaken. Most of the exceptions can be satisfactorily reconciled with the distinction if it is conceived and applied in the manner suggested in the paper. (shrink)