The purpose of this paper is to investigate one of the basic assumptions of the Szondi test, namely, that in selecting pictures in the test subjects tend to assign consistent meanings to the pictures, although these meanings are not explicitly stated. In the following experiments, this assumption was tested by obtaining explicit reactions to the pictures in the form of identifications, using two different experimental methods. Similar studies have been made by Klopfer and Borstelmann (1950), Rabin (1950a), Fosberg (1951), Davis and Raimy (1952), and by Dudek and Patterson (1952). In all these investigations, American subjects with some experience of psychology or psychiatry were used. The most extensive work was done by Klopfer and Borstelmann, who used a sample of American psychology students, and suggested that their methods might be employed on “other segments of the population”. In the following experiments, their methods have been largely adopted for English subjects, of varying ages and occupations, including students.