which we consider as attributes are often considered as independent objects. The best-known case of this kind, one to which I have referred incidentally before, is that of sickness. While we consider sickness as a condition of an organism, it is believed by primitive man, and even by many members of our own society, to be an object which may enter the body, and which may be removed. This is exemplified by the numerous cases in which a disease is extracted from the body by sucking or by other processes, in the belief that it may be thrown into people, or that it may be enclosed in wood in order to prevent its return. Other qualities are treated in the same way. Thus the condition of hunger, exhaustion, and similar bodily feelings, are considered by certain primitive tribes as independent objects which affect the body. Even life is believed to be a material object that may become separated from the body. The luminosity of the sun is considered as an object that the Sun himself may put on or lay aside.
I have indicated before that the concept of anthropomorphism seems to be one of the important categories underlying primitive thought. It would seem that the power of motion of the self and the power of motion of an object have led to the inclusion of man and movable objects in the same category, with the consequent imputation of human qualities to the moving objective world.
While in many cases we can see with a fair degree of clearness the fundamental concepts underlying these categories, in other cases these are not by any means clear. Thus the concept of incest groups—those groups in which intermarriage is strictly forbidden—is omnipresent. But no satisfactory explanation has so far been given for the tendency to combine certain degrees of blood relationship under this view-point.
Much material for this field of inquiry is contained in the works on comparative anthropology, but I believe a more thorough psychological analysis of the accumulated data may reveal important new information.
We will now turn to the consideration of another group of psychological phenomena that seem to me of considerable importance. In all forms of society certain groups of activities and of thoughts appear in certain typical associations. Thus in our modern society the consideration of cosmic phenomena is constantly associated with the efforts to give adequate explanations for them, based on the principle of causality. In primitive society the consideration of the same phenomena leads to a number of typical associations which differ from our own, but which occur with remarkable regularity among tribes living in the most remote parts of the world. An excellent instance of this kind is the regular association of obser-
