This paper investigates one of the great achievements of twentieth-century cell biology: determining the structure of the cell membrane. This case differs in important ways from the better-known case of the identification of the DNA double helix as the carrier of genetic information, especially regarding the evaluation of potential evidence in light of prior theoretical commitments. Whereas it has been argued that adherence to a structural hypothesis enabled Watson and Crick to ignore a surplus of (potentially confusing) empirical findings, similar adherence to an elegant and universal structural hypothesis, we argue, unduly shielded the so-called "unit-membrane' model from legitimate challenges on the basis of known phenomena.