COLUMBIA OBSTRUCTION METHOD pie) and, with equally easy access to both, supposedly goes to the one having the highest incentive value at the moment. It is difficult to say who first applied this type of test, but certainly it has been used extensively by various workers on lower forms in approaching the problem of the relative potency of tropisms, and of instincts under different conditions. The literature of the last several decades contains so many studies of this sort that it is hardly necessary to cite specific investigations. Mast has perhaps used the method more widely than any other. More recently (1924) Moss ( 8) compared the food-sex drives in the white rat by this method, using a simple technique, which was greatly improved by Tsai (18) the following year. Jenkins (Part I, 1, last 4 pages) developed a standardized apparatus for testing animals by this method in 1925, but we have made no use of this device in our research work on motivation at Columbia, because we have felt that the method, even as improved by Jenkins, is of slight value in this field. In describing the latter apparatus, the writer (4) (Part I, 1, last 4 pages) pointed out the more obvious and fundamental defects of any form of Choice Method, as a means of analyzing dynamic behavior. It is quite impossible to isolate a single drive, since the method itself requires that two physiological states, corresponding to the two incentives used (food-sex, for example) be simultaneously aroused in the organism. But the two states induced are likely to be interdependent and the corresponding drives, unless they are genuinely antagonistic, thus influence one another in various manners and degrees instead of operating as distinct internal tendencies. Furthermore, it is impossible to say in our present state of knowledge precisely how much starvation should be balanced against a given amount of sex deprivation in such a case. The Method of Choice appears to be ill suited to quantitative, analytical work on motivation.The Learning Method (or Training Method) of studying motivation has been employed with some success within a fairly limited field. This method is an adaptation of the usual learning experiment in that different incentives are used in the setting up of the same habit, and the various incentive-drive situations ranked in terms of some common index of learning efficiency (speed, accuracy, etc.). This method has been used in a score or more of studies COLUMBIA OBSTRUCTION METHOD 7 beginning with the early work of Yerkes (1907) on punishment vs. reward; examples of the application of the method to certain 8 COLUMBIA OBSTRUCTION METHOD tinguished from a training method of measuring motivation. The basic principle employed in all these attempts is the general one involved in the usual learning experiment, which has come down to us from Lubbock. As early as 1882, Lubbock (6) laid down the principle that the best method of testing an animal as to its intelligence, etc., is to " interpose some obstacle" between the animals to be tested and some d...