Our evidence so far as it goes points to some innate basis of difference. This basis does not seem to be the level of general intelligence or of emotional stability, nor in any general superiority or inferiority in learning or attention, but in such specific factors as greater speed of reaction, ease of breaking habits, readiness to make snap judgments, and independence in the face of majority influence. The last of these differences is the one most clearly indicated.... If one man is by nature more keyed up for speed and flexibility, and the other is more designed for regularity of function, we can hardly expect that government of the hyperkinetic by the phlegmatic and for the phlegmatic can fail to develop periods of stress and strain.[22] Floyd Allport classified students into typical and atypical members of the community, discovering whether they took up prevalent or minority positions on various political questions. He then applied a battery of tests to the students and found that certain traits of the atypical (whether the opinions were “radical” or “conservative”) proved to be homogeneous. Thus, not only do the “extremes” meet, but also those who hold minority positions along a hypothetical continuum of opinion distribution.[23] W. H. Cowley, at the instigation of Charles E. Merriam and L. L. Thurstone, undertook to compare leaders in chosen situations. He used twenty-eight tests of such traits as aggressiveness, self-confidence, intelligence, emotional stability, and speed of reaction. While his tests were differentiating within the same group between leaders and followers, he reports that they did not distinguish between leaders as a class and followers as a class. He thus felt justified in denying that leadership is a universal trait of particular persons, and criticized efforts to itemize “traits of leadership.”[24] Further applications of the blanket-test technique are under way.[25] Gilbert J. Rich opened out a new field of investigation when he studied certain complex physiological and biochemical variables among leaders and followers. He measured the hydrogen-ion concentration of the saliva, the acidity of the urine, alkali reserve of the blood, and creatinine excretion of the urine. The least excitable subjects showed the most acid saliva and the most acid urine. The rating of the subjects was open to criticism, and was obviously much less refined than the biochemical techniques employed.[26]