Using flat drawings, intra-reversal times were obtained for two kinds of fluctuations, perspective reversals in a figure eliciting apparent depth and reversals of lateral organization in a figure not eliciting apparent depth. The durations of the alternate percepts (P1 and P2) between reversals early in a viewing session were compared with recently published data of the same type. In the latter study, P1 was found to be of longer duration than P2, both with a flat drawing and a rotating skeletal cube. These prior data were explained in terms of satiation theory; however, we found no significant differences between P1 and P2 with two different groups of Os and with two figures. Explanations for these differences in results were advanced. The use of grouped percept-duration data to support steady state (e.g., satiation) theories of perceptual fluctuation phenomena was criticized.

Four Os drew maps of their autokinetic movement for a central light when it was the only stimulus, and when another light was adjacent to it. Eight directions (at 45° intervals) at each of two distances from the central light (1.27 and 2.54 cm) were used to yield 16 different placements of the light-pairs. The addition of the second light in any placement resulted in a significant reduction in the amount of movement and an increase in its latency. At either of the distances used, both lights were still viewed in the fovea, and the results did not differ in this respect. The direction of the second light from the central one did exert a significant influence, however. The results are compatible with the view that autokinesis results from a combination of eye movements and efferent tension.

Some earlier methodological and theoretical criticisms of studies of revecsible perspective reported by Price have recently been examined by him. Several of these criticisms are expanded here, and some problems inherent in the use of grouped data relative to a satiation theory of perceptual fluctuation such as his are discussed.Geometric "ambiguous" figures, such as the Necker cube, yield two or more percepts which alternate as sudden changes in perspective. Some authors believe that as a given percept persists, instability increases as a result of accumulating inhibition, fatigue, or satiation forces in the brain and that at some point in time the percept changes as a result (McDougall, 1906;Koehler & Wallach, 1944). Price ( 1967b) recently proposed a two-process satiation theory,? and we replied (Sadler 8: Mefferd, 1970) that grouped daca do not provide appropriate support for a satiation-based explanation of perspective fluctuation. W e also questioned certain methodological procedures used by Price, to which he has responded (Price, 1971). Since these issues are critical to his conclusions, we expand our earlier discussion here.Over the years our results with a large variety of so-called ambiguous stirnuli are "seen" in alternating percepts, regardless of the relative dominance of the separate percepcs or of the specific instructions, and always involve very large intra-individual differences. For a large proportion of our 0s such variation with diverse viewing conditions, instructions, and stimuli is evident in daca pertaining to the first percept seen ( P l ) in different sessions, mean rates of fluctuation during different sessions, and most importantly for the present discussion, in the duration of succeeding percepcs within a session. This variation of intervals between flucmations is extremely large for most Os-a given percept (either PI or P2) may persist for a second or so, then be followed by one that persists for many seconds, then by one which lasts for less than a second, and so on.This type of intra-individual variation occurs at all stages in a person's ex-'and Baylor College of Medicine, Houston, Texas, and the University of Houston.Trice is privileged to contend that an intervening variable type definition of satiation consrimtes a departure from the concepts of Koehler and Wallach (1944); however, those authors were the first to formulate explicitly the satiation concepts which Price is atrempting to extend. Also, we contend that satiation theory is basically the same whether it is given a neurophysiological base or whether the box is painted "incervening-var~nble black" -there is a brain in there somewhere! W e should say at this point that we generalized too far in stating that Price's curves of P1 and P2 became equal. W e agree thnc the curves of mean data show that P 1 is ahuays longer than P2. W e only meant to say that the dramatic initial differences between P1 and P2 were gone by the fourth or so fluctuatioo.

The SNS response decrement obtained by Mefferd and Wieland (1965) when a congitive task was imposed during a painful stimulus was reproduced with 15 Ss under conditions involving minimal motor activity. The results support the conclusion that cognitive activity/»er se was the major determinant of the decrement.After a decrease caused by alerting the S for the first passage no further significant changes in BSR occurred. The mean GSR (sum of transient decreases in resistance adjusted to a per minute basis) increased from .99 kilohms during the baseline period to 3.78 kilohms during the passage alone (two-tailed t test, p < 0.01) and to 6.60 kilohms during the cold pressor. Immediately after the passage was imposed on the cold pressor the GSR decreased, and the mean for the period was 4.00 kilohms-a significant decrease from the cold pressor period (p < 0.01, two-tailed t test). These results are the same as those obtained in our earlier study. They support the conclusion that the key determinant of the response decrement was not the motor activity of work association, and demonstrate that the results can be generalized to a larger population.

The interaction of three pairs of heteromodal sensory stimuli-^reading vs. continuous banging noise, continuous flashing light vs. banging noise, single flash vs. bang-were evaluated. Order of presentation was confounded both within (with multiple trials) and between subjects. Interstimulus time periods were 15 sec with the first two stimulus pairs, and varied from 15 to 0 sec with the last. Although most variables had large mean differences in the expected direction, the largest and most often significant effects were with GSR. In all cases, the responses were less with a stimulus imposed either simultaneously or within a short period (up to 15 sec at least) after another, than they were to the stimulus imposed alone. Habituation effects were marked.