The hippocampal formation has long been the focus of intense interest among neuroscientists. The discovery by Scoville and Milner (1957) that the mesial temporal lobe, including the hippocampal formation, played a central role in memory function in humans can be said to have started the modern era of research on this brain system. There followed two decades of animal research aimed at creating experimental models of the memory defects observed in humans, including my dissertation research, which met with scant success (Nadel, 1968). Indeed, when John O'Keefe and I started considering the hippocampus and its possible role in spatial mapping several years later, one of our greatest challenges was to analyze the extant lesion literature in terms of this theory with the minimum amount of special pleading. There were already several hundred such studies, varying extensively in terms of exact size and regional location of the lesion, the means by which the lesion had been created, the kind of animal used, the nature of the task, the kind of motivation employed, and so on. Yet, it was our feeling in 1976, when we terminated the review of the lesion literature, which was ultimately published in The Hippocumpiis CIS a Cognitive Map (O'Keefe and Nadel, 1978), that the spatial hypothesis successfully handled the vast preponderance of these lesion studies as well as what was then known about the physiology of the hippocampal formation.In the years following publication of the book, most published lesion studies were concerned in some way with the spatial map hypothesis; most often, comparisons with rival hypotheses were involved. This was a frustrating period because, although many thought they were testing our ideas, too frequently the major themes we hoped to stress were misconstrued. In this article I dig back into the book we published in 1978 and reiterate some of the themes that were important to us then, and remain so now. By so doing, 1 hope to bring out some of the reasons why we thought, and continue to think, that space is so special as to warrant a structure as compelling as the hippocampal formation.I will concentrate on three major issues. First, I will discuss at some length the fact that our theory was about a ~p a t i a l memory system, not simply a "spatial" system. Second, Iwill discuss our assumption that the spatial memory system we allocated to the hippocampus was but one of many spatial systems, and that the role of these various systems can best be understood if one takes into consideration an animal's behavior in its natural habitat. Third, I will briefly discuss our view that in the human hippocampus, at least in the left hemisphere, the mapping system represents something more abstract than physical space.Before turning to these substantive points, however, a brief digression into methods and first principles is in order. Some years before the book was published, O'Keefe and I published a paper (Nadel and O'Keefe, 1974) in which we presented some of the reasons for hypothesizing that the hippocamp...