It may be reasonably argued that of all classes of known cellular proteins, the ion channels are the least understood biochemically. Of the large number of channel proteins known from cellular electrical behavior to exist in the mem­ branes of higher organisms, only three-the nicotinic acetylcholine receptor, the Na+ channel, and the mitochondrial porin channel-have been obtained in a relatively purified state (28). Even in these cases, serious questions remain about the physiologically "correct" functioning of these purified proteins. A single peculiarity of ion channel proteins explains this unhappy situation: the lack of a suitable functional assay for integral membrane channels. Seeking to purify an enzyme, for instance, a biochemist can straightforwardly develop a specific assay to pursue and eventually to apprehend the protein with a battery of purification methods. But the only criterion of functionality of an ion channel is its ability to make a membrane "leaky" to specific ions. To be sure, this "leak" may be turned on and off in very specific ways and may be modified pharmacologically, but the fact remains that ion channels, by their very definition, merely catalyze the passive flow of ions down their thermodynamic gradients. It is therefore necessary, in attacking purified channel proteins, to develop a system in which the flow of ions across a membrane may be detected. Histor­ ically, two such model membrane systems have been employed: planar bilayers and liposomes (10, 22, 26, 27, 32-36). The planar bilayer system has been largely favored for ion channel work because of its obvious advantages: the ability to voltage-clamp the model membrane and to measure single channel current fluctuations, and the easy accessibility to both aqueous phases. But there are two serious disadvantages in the use of planar bilayers in biochemical