outside his own university have the most effect on a scientist's chances for recognition. Although the data lend themselves to various interpretations, they do not permit a conclusive explanation of the high correlations between academic setting and both productivity and recognition. On the one hand, these findings favor an elitist view of scientific activity. Price has shown that the majority of scientific work is performed by a relatively small number of scientists,50 and these data indicate that the most talented students trained at the best graduate schools by the most eminent teachers are most likely to become and to remain highly productive. On the other hand, the data may reflect a pattern of scientific activity that is rapidly changing. In the past 10 or 15 years, federal funds for scientific research have become available to an increasing number of universities, augmenting the amount and presumably the quality of research conducted in these institutions. As these schools become more strongly committed to and involved in research programs, the productivity and recognition of their faculties may approach the pattern now characteristic of the major universities. 50 Price, loc. cit.