IN 1974, TWO ARTICLES APPEARED in this Journal which are importantly, albeit not obviously, related. The present essay is intended to apply the observation of Barr Rosenberg and Michel Houglet [7]-that popular data bases have nontrivial errors -to the work of Richard McEnally [5] dealing with the return behavior of high risk common stocks. This paper is organized as follows: The next section contains some of the results of a replication of the McEnally study and emphasizes some important characteristics of the data that were not fully reported by him. In part III, the Rosenberg and Houglet study is reviewed and the results applied to the McEnally work. Next, the impacts of legitimate outliers are examined and finally, some general conclusions are drawn from the analyses.