The solution of the geodetic boundary value problem for determining the anomalous potential from gravity measurements in the spherical approximation, taking into account the relief and compression, was developed in sufficient detail in 1960 and is also based on the results of G. G. Stokes. Flattening of the reference surface was taken into account by D. V. Zagrebin in several works dated 1940–1970; in 1956 M. S. Molodensky proposed a simpler method for an oblate ellipsoid, based on the already known one for a sphere. It turned out that at the very beginning of developing the mentioned solution, an error was made when obtaining the derivative of the reciprocal distance between two points on the ellipsoid in the direction of the outward normal; later it was repeated in issue 131 of the Proceedings of the TsNIIGAiK. That inaccuracy did not influence the conclusions of V. V. Brovar in 1996 and 1999, where the relief was added to the consideration of compression. This solution of Molodensky’s problem, taking the compression into account, was seldom used, and that is probably why the mistake went unnoticed. V. V. Popadyev saw it and found the correct expression for the mentioned derivative, I. I. Mezhenova adapted subsequent conclusions for the gravity disturbances.