Newcomb's (1898) Tables of Mars with Ross's corrections have been used to compute its ephemeris in national ephemerides. Clemence (1949 and 1961) constructed a more precise theory of Mars and determined preliminary mean elements using a limited number of observations. Laubscher (1971) compared Clemence's theory with about 11000 observations by meridian circles from 1851 through 1969 and about 800 observations by radars and derived an improved set of mean elements. Duncombe and Kubo (1977) compared the ephemeris of Mars, in which Clemence's theory and Laubscher's elements were adopted, with the observations made by the meridian circle at Washington in 1972 and found large systematic residuals, which were larger than residuals by Newcomb's theory. Therefore, Duncombe and Kubo suspected that Laubscher's treatment of data was erroneous. On the other hand we had undertaken a reconstruction of Clemence's theory. In the meantime Kubo and Seidelman independently have found Laubscher's error in the determination of the inclination.