It is interesting to note that even though the transition state associated with the postulated mechanism has the composition { Cr(CN)Hg4+ ]* the rate was found to be independent of [Hg2+]. This situation arises because the laboratory rate expression is based on variations for formal concentrations rather than concentrations of predominant species. In the present situation CC~CN S [Cr(CN)Hg4+] so that the rate expression, eq 10, when written in terms of the predominant species, does correctly reflect the composition of the transition state.This system can be contrasted to the Hg2+ "catalysis" of the aquation of CrC12+,21 where the rate varied as the first power of [Hg2+]. In that case no evidence was obtained for a cation association complex.Structure of the Cation-Cation Complex.-The association of two cations as in eq 2 is a very unusual type of reaction, especially an association as stable as the one encountered here. Chromium(II1) and neptunyl-(VI) ions are known to associate in a rather weak but kinetically stable complex. 37 hlercury(I1) is known to form rather stable complexes with Co(NHJ5CN2 +, 58 and the ion CrNCSHg4f is formed in the reaction of (H20)sCrNCS2+ n-ith Hg2+.39 Other AI-thiocyanate and M-azide association reactions with Hg2+ have been reported. 40 a dThe reaction of CrNCS2+ and Hg2+ (eq 20) stands in contrast to that occurring between CrSCN2+ and Hg2+ (eq 21). Since mercury(II), a soft acid, will co-(H20)&rNCS2+ f HgaU2+ = (H20)jCr-SCS-Hg,,,'i+ (20) (H20)&SCSz+ + Hgaa2+ = Cr(H20)63C + Hg-SCX-'. (21) (37) J. C. Sullivan, J . A m . (41) D. A. Loeliger and H. Taube, iiiorg. Chem., 5, 13i6 (1966). 315 (1964). (1959).ordinate to the softer base site of an ambidentate ligand, the high stability of the cyano dinuclear species is surprising. A species of the structure (H20)jCr-CNwould seem to stand little chance of being a stable one, as nitrogen donors generally do not provide especially stable complexes with mercury(I1).We propose that a structural change accompanies formation of the dinuclear complex, as represented by (H20)6CrCNZ+ f HgaQz+ = (H20)5Cr-NC-Hga+ (2') The proposed internal isomerization of cyanide finds support in the spectral shifts accompanying this reaction [A,,,( CrCN2 +) 5250 -4, X, m, , ( Cr-NC-Hg4 f) 5650 A], considering that the nitrogen end of cyanide has a smaller ligand field strength than the carbon end.42The spectra of the "supercomplexes" (between (NH3)5-Co-CN2+ and Hg2+) found by SieberP suggest that a similar rearrangement to form (NHs)jCo-NC-Hg4+ has taken place [X,,,(Co(NH3)5CN2+) 4400 A, A, , , (Hg2+ adduct) 4480 -41. Support for the postulated internal isomerization comes from our o b~e r v a t i o n~~ that the same species, Cr-NC-Hg", arises from the reaction of mercury(I1) with the isocyano complex, (H20)5CrNC2+, the latter prepared from the reaction of (NH3)6CoCN2+ with chromium(II).3 A precedent for the proposed linkage isomerization accompanying dinuclear complex formation in eq 2' can be cited. The main products of the reaction of (H20)Cr-SCN2+ and Hg2+ were ...