Thompson's (1979) comment points out four areas of difficulty in interpreting my results (Ksir, 1978). The first, that strain differences in TI duration might have masked the effects of scopolamine, is a possible explanation of the difference in results between our two studies. However, the saline and methylscopolamine durations reported in my paper were quite similar to the durations reported for their birds under similar drug conditions. Thus, strain differences in scopolamine action, rather than in baseline TI duration, would have to be invoked to explain the different results. I do not know of studies reporting strain differences in scopolamine action among chickens.Their fourth point, that I referred to activity effects as changes in "open-field" activity when both experiments were done in fairly small activity chambers points out a careless choice of words on my part, but since there were no substantive differences in the results of our two experiments, this is a minor point.Their third point, that I used an "unusual" testing procedure in my second experiment, is irrelevant. In my paper, under the bold heading "EXPERI-MENT 1," is the statement, "Experiment 1 was an attempt to replicate as nearly as possible .... " The second experiment had a different purpose entirely, that of separating induction effects from duration of TI once induced. That this analytic procedure had not been published before is true, and in that sense it was "unusual."The most substantive failure of my experiment to replicate theirs was apparently my failure to use three successive inductions rather than a single induction . I had wished to avoid the possible difficulty of nonindependence of repeated tests with the same bird within a short period of time. I used 12 chickens per group rather than their 10, hoping to offset somewhat the loss of statistical power. The Thompson et aI. (1974) report had stated that "examination of the data indicated that there were no consistent changes across trials within a session, so the mean duration of TI for each bird for each session was used for analysis." This pooling no doubt had the effect of reducing variance. Thompson's letter states that analysis of the data from the first trial only showed no significant effect of scopolamine, which he says is similar to my result. However, if their data showed no consistent changes across trials, then we must assume that data from the first trial looked like the mean data, that is, there would have appeared to be a dose-related decrease in TI duration, but the variability of the first-trial-only data resulted in nonsignificance. Had I actually obtained such results, that is, data that sort of looked like theirs but didn't reach significance due to variability, I would not have submitted them for publication. The reason for publishing my results was that the means for the drug and control groups did not vary much and were remarkably consistent (and similar to their control values). In other words, one can fail to find an effect either because the effect ...