A RECENT paper by Dr. Harold Jeffreys (Mon. Not. Eoy. Astro. Soc., 97, 1; 1936) deals with some well-known articles by de Sitter (Bull. Astr. Soc. Neth., 2, 55; 4, 129. Also K. Akad. Wetens. Amst., 17). De Sitter assumed hydrostatic stress in the earth's interior, and computed the ellipticity and related constants from the precessional constant. On comparing the results with the motion of the moon's node and perigee, the greater part of which can be explained by Brown's theory, there is a balance depending upon the figures of the earth and moon. The former was taken by de Sitter to be calculable from his theory, and the small balance he attributed to the moon's ellipticities. As Jeffreys points out, however, the assumption of hydrostatic stress in the interior does not agree with other data. For example, accepting his solution, the probable error of the difference between gravity at the poles and at the equator is 1 milligal (0-001 cm./sec.2), but Heiskanen's ellipticity of the equator alone gives a variation of 38 milligals at the extremes. In addition, the relativity correction was applied by de Sitter with the wrong sign, a fact which he pointed out to Jeffreys about five years ago, and this produced errors in some of the constants.