The present inquiry was started with a much more specific aim than that which it later developed. In October 1928 was published a startling announcement of a very high relationship between reflex speed and intelligence. This result was so strange that it was taken by many psychologists "cum grano salis"; but it was obviously of surpassing interest if true. Then must the psychology and measurement of intelligence have been revolutionized. Here was the realization of the intelligence tester's dream - a single physiological measure obtainable very early from the subject and independent of environment. Here was "g". The figures quoted appeared strikingly significant, r = .87 ±.04 from two groups of over 40 subjects. Such a result must be thoroughly tested before it would be time to interpret it with its probably far - reaching effects. With this end in view, work was started. The expense of the method used by the American authors was prohibitive so a much cruder technique had to suffice. With this method no relationship was found. The technique was far too crude, however, to be trustworthy and an improvement was sought. Before I had succeeded with a better method, a. second article appeared from the chief author denying completely the previous finding and explaining this as based on unreliable data. Since the data this time seemed to have been subjected to rigorous tests, the finding Of no relationship was accepted and other more general problems were then considered. That further work was necessary upon some of these seemed apparent, and it is hoped that the results obtained will be of some interest and value.