1 4 V O L . 3 , D O C . 1 0 a B O L T Z M A N N ’ S P R I N C I P L E approximation, it nevertheless fails because he assumed that density fluctuations in different volume elements are statistically independent and because he did not take into account the fact that the cor- relation function diverges as one approaches the critical point, where density fluctuations in different volume elements become correlated. This was pointed out a few years later in Ornstein and Zernike 1915. For further historical discussion, see Vol. 3, the editorial note, “Einstein on Critical Opales- cence,” pp. 283–285. [12]For quantitative considerations, see Einstein 1914a (Vol. 3, Doc. 26).