We must bear in mind a fact which is familiar enough, though it has been too often forgotten by legislators and criminalists. Society is not a homogeneous aggregate, but on the contrary an organism, like every animal organism, composed of tissues of varying structure and sensibility. Every society, in fact, with its progressive and increasingly distinctive needs and occupations, is a product of the union of social classes which differ greatly in their organic and psychical characteristics. The physical constitution, the habits, sentiments, ideas, and tendencies of one social stratum are far from being the same as those of other strata. Here again we have, as Spencer would say, the law of evolution through a departure from the homogeneous to the heterogeneous, from the simple to the complex, or, in the words of Ardigo, a natural formation by successive distinctions. Amongst savage tribes this distinction of the social strata does not exist, or it is far less marked than in barbarian societies, and still less than in civilised societies. Every schoolmaster with a bent for psychological observation separates his pupils into three classes. There is the class of industrious pupils of good disposition, who work of their own accord, without <p 94>calling for strict discipline; that of the ignorant and idle (degenerate and of weak nervous force) from whom neither mildness nor severity can obtain anything worth having; and that of the pupils who are neither wholly industrious nor wholly idle, and for whom a discipline based on psychological laws may be genuinely useful. This is the case with large bodies of soldiers or of prisoners, for all associations of men, and for society as a whole. These partial organisms, due to the constant relationships of a life more or less in common, are in this respect reproductions of society as a whole, just as a fragment of crystal reproduces the characteristics of the unbroken crystal.[13] [13] There is, however, some difference between the manifestation of the activity of a group of men and that of the aggregate society. Between psychology which studies the individual, and sociology which studies the society, I think there is room for a _collective psychology_, to study more or less defined groups. The phenomena of these groups are analogous, but not identical with those of the sociological body properly so called, according as the union is more or less definite. Collective psychology has its field of observation in all unions, however occasional, such as the public street, the markets, workshops, theatres meetings, assemblies, colleges, schools, barracks, prisons, and so forth. Many practical applications of the data of collective psychology might be given. An example will be found in a future chapter, when I come to consider the psychology of the jury.