social systems are the fruit of design, and that they were deliberately devised to ensure the results which they unquestionably achieve. The latter was the opinion of Fison and Howitt, and it is the opinion of their distinguished friends and disciples, Spencer and Gillen.
In the broadest outline, omitting details and minor differences, an aboriginal Australian tribe is divided into two, four, or eight exogamous classes; that is, it consists of two, four, or eight divisions with a rule that no man may marry a woman of his own division, but may only take a wife from a single one of the other divisions. Thus, if the tribe is divided into two exogamous classes, a man is forbidden to choose his wife from among, roughly speaking, one-half of all the women of the tribe; if the tribe is divided into four exogamous classes, then three-fourths of the women are forbidden to him; and if the tribe is divided into eight exogamous classes, then no less than seven-eighths of the women of the tribe are forbidden to him. So strictly are these rules enforced that in the old days breaches of them were commonly punished by putting both the culprits to death.
With regard to descent, when a tribe is divided into two exogamous classes, the children are always born into the class either of their father or of their mother, the custom in this respect varying in different tribes; for in some tribes the children always belong to their father's class, and in others they always belong to their mother's. When a tribe is divided into four or eight exogamous classes, the children are born into the class neither of their father nor of their mother, but always into another class, which is, however, determined for them without variation by the particular classes to which their parents belong.
It will hardly be denied that these systems, particularly the rule of the four-class or eight-class organisation, that children can never belong to the class either of their father or of their mother, have at least a superficial appear-
