The principal results of the Vanguard 3 magnetic field experiment are summarized in a report and a paper [Cain et al., 1962 a, b]. Since some of the material presented in these documents has been used by several authors, it is worth while to clarify a few points and to report a modification indicated by a later analysis. The Vanguard 3 report gives a series of 63 spherical harmonic coefficients fitted to the data themselves by the iterative procedure of Jensen and Cain [1960] (reviewed by Heppner [1964]). This field, designated Vanga, used g• and h• • to n --m --7 and gave an rms deviation from the data of 21 7-The data selected for this fit were the 2797 observations given by Cain et al. [1962a] as L --0 (--best data), excluding four such observations with residuals (measured F-computed F) over 100 7. These observations were those recorded at Blossom Point, Maryland, on October 4, 1959, at Oh 47m 23s, Oh 47m 37s, 19h 50m 11s and 19h 50m 29s. In the Vanguard 3 report we