tn a recent article in this journal, Niemi and Weisberg (1974) have compellingly shown the falsity of the plausible assumption (made by Grofman, 1968 and others: see citations, Niemi and Weisberg, 1974) that Guttman scalability (see Stouffer, et al., 1950 MacRae, 1970) is equivalent to single peakedness (see Black, 1958 Arrow, 1962, and Coombs, 1964). They have further demonstrated that, in general, information about individual votes on a series of items found to be Guttman scalable does not yield the locations of those individuals' ideal points in some space, unidimensional or otherwise. Niemi and Weisberg (1974) focus on the question of whether Guttman scalability implies single-peaked preferences. If this implication would be demonstrated then various theorems on single-peakedness and equilibria in voting outcomes (e.g., Black, 1958 Arrow, 1962) would have been shown to be applicable to the legislative or judicial context where sizeable subsets of Guttman scalable issues have been found (MacRae, 1970; Rhode, 1972). Furthermore, spatial models of voter choice which postulate single-peaked preferences (e.g., Downs, 1957) could be applied to data on attitude dimensions shown to be Guttman scalable (e.g.,