MOLECULAR FORCES 265 1806). Einstein wrote Ostwald that his work on this topic had been stimulated by Ostwald 1891 (see Doc. 92). In Einstein's theory, the force between two molecules is derived from a potential: P = P"-c c'0(r), where c and c' are constants characteristic of the two molecules, and /(r) is a universal function of the distance between them. Einstein was intrigued by the possibility of an "innere Verwandtschaft" between this force and gravitation (see Docs. 100, 101). He continued to work on capillarity after publication of his first paper, but apparently without success (see Doc. 119). By the spring of 1901, Einstein had applied his theory of molecular forces to salt solutions. He studied infinitely dilute solutions in order to avoid the complications introduced by interactions between solute molecules (see Doc. 101). The results of this investigation were published in Einstein 1902a. By applying the theory to mixtures of neutral liquids, he drew the conclusion that, for the molecules of such a liquid, c should be proportional to the specific volume. If this result were verified experimentally, he wrote, it would mean the end of the mo- lecular-kinetic theory of liquids (see Doc. 127). Einstein also attempted to extend his theory of molecular forces beyond liquids he planned to use the results as the basis of a doctoral dissertation (see Doc. 100). In September 1900, after studying Boltzmann's work on gas theory, he stated that "die hypothetischen Kräfte" between molecules play no essential role in the kinetic theory of gases (see Docs. 75, 76). The following year he realized that these forces play a role in transport phenomena in gases. He set out to evaluate the coefficients of diffusion, heat conduction, and internal friction in terms of his constants c (see Doc. 101). In this connection, Einstein was troubled by the need to take into account a finite molecular size in addition to the central force law between molecules (see Doc. 102).  There is reason to believe that Minkow- ski was a more important influence. He gave a lecture on capillarity during Einstein's last term at the ETH. Einstein is said to have re- marked, "Das ist die erste Vorlesung über mathematische Physik, die wir am Poly hören!" (Cited by Louis Kollros in Seelig 1956, p. 21.)  Ostwald attributes the idea of such a re- lationship to van't Hoff (see Ostwald 1891, p. 1142). A similar relationship is suggested in a popular-scientific book (Bernstein 1870, pp. 142-143) that Einstein read in his youth (see Einstein 1979, pp. 12, 14).  This is probably a reference to the van der Waals theory (see Doc. 127, note 16).  See Boltzmann 1896, §§12, 13. Following Maxwell's analysis, Boltzmann showed that the difficult mathematical problems connected with the study of these phenomena could be solved for one particular force law (inverse fifth power).  The van der Waals theory of gases and liquids is based on the assumption that mole- cules have a finite extension as well as an attractive central force between them (see the first two sections of Boltzmann 1898). The failure to introduce a characteristic length for each molecule (molecular "radius") is the fun- damental difficulty in Einstein's theory of molecular forces (see the editorial note pre- ceding Einstein 1901 in Volume 2).