site is vacant, all six O(3) oxygens have shifted 0.50 A toward Ni(1) and the center of the double six ring-a striking demonstration of the ability of the aluminosilicate framework to adjust to change: in cation position. The Ni(l)-0(3) distance, 2.29 A, is an average value that includes O(3) oxygen positions for SI sites containing no nickel ions. This fact is illustrated by the O(3) thermal ellipsoids, which are elongated in the Ni-0 direction. Finding the nickel ion on both sides of the supercage six ring may be due to the presence of the residual water, O(w2), and the resultant Ni(3)-O(w2) bond formation.Although the nickel ions prefer SI sites, they are distributed among four, and possibly five, different sites, with no site completely filled. This distribution agrees with conclusions drawn by Barry and Lay.a From esr studies of manganese in mixed cation X zeo-lites, these authors found manganese(I1) distributed in a number of sites and concluded that Mn2+ does not have very strong site preference.Also, from ir studies of CO adsorbed on various cationic forms of zeolite Y, Angel1 and Schaffer4 found that Ni2+ and Go2+ ions have no overwhelming preference for site I. Although site I has the highest occupancy in nickel faujasite (66%), it is not fully occupied. I n this respect, ie., a lack of high site preference, transition metal zeolites differ from calcium in zeolite Y, which prefers site SI,24 and rare earth ions in zeolites X and Y, which prefer SI' sites.l1Vz6J6 (24) R. P. Dodge, unpublished research.The purpose of the work described in this paper was to investigate the dielectric behavior of the protein bovine serum albumin. The investigation was made to try and solve some of the problems that had arisen from measurements obtained by other workers in the radiowave and microwave frequency range. To do this, measurements were made over as wide a concentration, temperature, and pH range as possible. It was confirmed that there was a subsidiary dispersion in the frequency range 200-2000 MHz, and the possible molecular interpretation of the dispersion is discussed. It can be concluded that this dispersion is probably due to water bound to the protein. By assuming a dielectric mixture theory the variation of the amount of bound water with concentration and pH is estimated.(1) In this paper, bovine Berum albumin will be abbreviated as BSA.