L may be. That is, the intrinsically strong Mo-Mo bond necessitates an intrinsically weak Mo-L bond. This "metal-metal trans effect" operates to some extent in the other dinuclear carboxylates. Thus, in Cr2-(0Ac)4(H20)2s and Cr30(0Ac)6(H20)3+,28 the average Cr-OAc distances of 2.018 (2) and 1.98 (1) A are comparable to each other and to the Cr-OH2 distance in the trimer, 2.02 (1) A, but all are significantly shorter than the Cr-OH2 distance in the dimer, 2.272 (3) A. In the dimer, the water molecule is trans to the Cr-Cr bond; in the trimer, to the Cr-0 bond. Again, the Rh-OH2 distance of 2.310 (3) A in Rh2(OAc)4(H20)2 is substantially greater than the average Rh-OAc distance of 2.039 (2) A. Nowhere is the trans effect more dramatically illustrated than in the molybdenum system, however; therefore, Mo(II) appears to have an intrinsically greater metal-metal bonding tendency in the carboxylate framework than do Cr(II) and Rh(II). This has long been suggested from theoretical considerations and the available X-ray data,7 but never so conclusively demonstrated. One expects, then, that if good crystals of unsolvated M2(02CCF3)4, M = Cr, Rh, could be obtained, which may not be possible, the Cr-Cr and Rh-Rh distances would definitely be longer than the Mo-Mo distance in Mo2(02CCF3)4, allowing for differences in covalent radii. Moreover, if crystalline M2(02CCF3)4(py)2, M = Cr, Rh, could be obtained, which should be possible, the metal-nitrogen distances should definitely be shorter than in Mo2-(02CCF3)4(py)2, again allowing for differences in covalent radii. The lengths found for metal-to-axial ligand bonds in the carboxylates thus provide valuable information about the implicit tendencies of the metals toward mutual attraction.