An analysis of the data of table 5 reveals five methods or reactive tendencies which appeared more or less definitely in the following order: (a) Choice of first box at the left, because of experience in problem 1. This tendency was very quickly suppressed by the requirements in connection with problem 2. Indeed one of the most significant differences which I have discovered between the behavior of the primates and that of other mammals is the time required for the suppression of such an acquired tendency. The monkey seems to learn almost immediately that it is not worth while to persist in a tendency which although previously profitable no longer yields satisfaction, whereas in the crow, pig, rat, and ring dove, the unprofitable mode of response tends to persist during a relatively large number of trials. (b) The tendency to choose, first, a box near the left end of the group, to go from that to the box at the extreme right end of the group, thence to the one next in order, which was, of course, the right box. This tendency appears fairly clearly from May 7th on. (c) The box at the extreme right was first chosen and then the one next to it. For example, in setting 2, box 4 would be chosen first, then box 3. Or, if this did not occur, the method previously described under (b) was likely to be employed, as for example, in setting 8, where such choices as 7.6.5.1.8 appear. (d) In certain series there appeared a marked preference for a particular box, usually box 3 (see results for May 24). This was doubtless due in a measure, if not wholly, to the fact that box 3 was the right box twice in each series of ten settings. But it should be added that the same is true of box 7, for which no preference was manifested at any time. (e) Direct choice of the right box. The five reactive methods or tendencies enumerated above roughly appeared in the order named, but there were certain irregularities and the order as well as the time of appearance varied somewhat from setting to setting. In general, method c was the most frequently used prior to the development of method e, the direct choice of the right box. TABLE 5 Results for Sobke, _P. rhesus_, in Problem 2