Prolonged viewing of a grating raises the threshold for detecting gratings of similar orientations (Gilinsky, 1968) and spatial frequencies (Mantle and Sekuler, 1968; Blakemore and Campbell, 1969). It has been suggested that adaptation is caused by prolonged inhibition of the detection channels and not by neural fatigue (Blakemore, Carpenter and Georgeson, 1971; Tolhurst, 1972; Blakemore, Muncey and Ridley, 1973; Kulikowski and King-Smith, 1973; Kulikowski, Abadi and King-Smith, 1973; Sharpe, 1974; Dealy and Tolhurst, 1974). Two arguments have been advanced by these authors for this view. The first maintains that the narrow bandwidths obtained with threshold summation of sine-wave gratings (Sachs, Nachmias and Robson, 1971; Kulikowski and King-Smith, 1973; Kulikowski et al., 1973) reflect the range over which channels can be excited. The bandwidths obtained with adaptation (Blakemore and Campbell, 1969) and masking (Campbell and Kulikowski, 1966; Stromeyer and Julesz, 1972). being considerably broader, are claimed to reflect the spread of inhibition between channels. Recent studies on threshold facilitation (Stromeyer and Klein, 1974; Nachmias and Weber, 1975; Barfield and Tolhurst, 1975), the detectability of frequency modulated gratings (Stromeyer and Klein, 1975), and probability summation (King-Smith and Kulikowski, 1975) suggest that gratings may not be detected by narrowband mechanisms. Thus, it is not clear that there is a discrepancy between bandwidths obtained with adaptation and other methods. A second argument that adaptation is caused by inhibition between channels and not by fatigue was presented by Dealy and Tolhurst (1974) who showed that when a 4.0 c/deg adapting grating just reached its visible threshold it started to raise the threshold of a subsequently presented, 6.7 c/deg test grating (3/4 octave higher than the adapting pattern). The authors