Now here we find a difficulty. According to our theory, the earliest attempts at the pictorial art should be pictures of men and women, but this is not, I believe, the case[38]. We must, however, remember that the idea of symbolic magic arises very early. This is natural. The representation of a thing is that thing in some degree. You have power over your representations, therefore you have power over the thing. The use of such power has an anti-social aspect, which forbids its common or public use except in the form, of a magico-religious ceremony. It is unlikely, therefore, that if such representations were made, they should have come down to us. Moreover, it is unnecessary that the magic object should bear any superficial resemblance to the thing it symbolises; indeed it is undesirable that it should be recognisable by others, since the practice for which it is destined is nefarious and illicit. An esoteric significance is enough. There is a very close connection between primitive art and religion. Thus the Palaeolithic drawings of animals in the dark caves of Périgord and Altamira, are undoubtedly connected with magico-religious ceremonies to give power over the beasts. For this reason then--the acquiring of a _prise_ over the object represented--we should hardly expect to find many early drawings of men and women, other than divinities. Even to-day many savages evince the greatest fear of having their likeness drawn. Nevertheless, these Neolithic drawings do exist, proving that there was no universal tabu on such representations. Moreover such drawings as those of the Bushmen show that primitive art at times uses drawings to record historical events, such as raids by other tribes. The comparative scarcity of primitive drawings is, however, easily explicable when we take the fact of magical beliefs into account. And there are sufficiently numerous examples of drawings of animals--bear, rhinoceros, lion, mammoth, bison, reindeer, to show that prehistoric man did have an intuition of his relation to other creatures. Furthermore, since the creative impulse does receive some, if unconscious, satisfaction in sex-relationship, expressed in word and action, there is the less need for technical expression in the early stages. We find at all events enough prehistoric drawings to show the recognition of relation, and the expressive activity, and these are the desiderata for an aesthetic fact. Leaving the most primitive level, we find the development of decoration. Pottery is shaped with some regard to form and symmetry, and simple ornament of a geometric character makes its appearance[39]. Much might be said on this subject, but we will confine ourselves to a few fundamental considerations.