This paper challenges the idea that memory is the special function of any single brain structure, an idea that developed from clinical cases of amnesia that had lesions in and around the hippocampus. There are many instances of amnesia and other evidence of memory functions in brain areas that do not involve the hippocampus. The evidence that medial temporal lobe lesions in animals produce uncontaminated memory deficits is reviewed and rejected, as well as the evidence that the memory task used in these experiments has any special relationship to the hippocampus. It is proposed that memory is not the function of any one structure, but is a part of local neuronal operations carried out in all cortical areas where the information to be remembered is processed and perceived. If this suggestion is correct, then a local lesion will cause a loss of local function and the memory for that function.