Resonance Raman data on bathorhodopsin (bovine and squid) at 95,77, and 4 degrees K support a mechanism of excitation proposed by Lewis in which both a protein conformational transition and chromophore structural alteration to a "dicisoid" configuration are required to generate the bathorhodopsin species observed in steady-state photostationary mixtures. However, these results also suggest that the molecular structure with a red-shifted chromophore absorption detected at room temperatures in 1 ps using picosecond absorption spectroscopy may not necessarily have the same chromophore conformation as the steady-state bathorhodopsin species.