This fact warns us to be cautious, but occasionally we have such evidence as is supplied by Europeans initiated into the mysteries of savage religion. Our best proof, however, of the existence of this exalted, usually neglected belief, is the coincidence of testimony, from that of the companions of Columbus, and the earliest traders visiting America, to that of Mr. A. W. Howitt, a _mystes_ of the Australian Eleusinia, or of the latest travellers among the Fangs, the remote Masai, and other scarcely 'contaminated' races.[4] If we can raise, at least, a case for consideration in favour of this non-utilitarian belief in a deity not approached with prayer or sacrifice, we also raise a presumption against the theory that gods were invented, in the despair of magic, as powers out of whom something useful could be got: powers with good things in their gift, things which men were ceasing to believe that they could obtain by their own magical machinery. The strong primal gods, unvexed by prayer, were not invented as recipients of prayer. To ignore this chapter of early religion, to dismiss it as a tissue of borrowed ideas--though its existence is attested by the first Europeans on the spot, and its originality is vouched for by the very absence of prayer, and by observers like Mr. A. W. Howitt, Miss Kingsley, and Sir A. B. Ellis, who proposed, but withdrew, a theory of 'loan-gods'--is not scientific. My own early readings in early religion did not bring rue acquainted with this chapter in the book of beliefs. When I first noticed an example of it, in the reports of the Benedictine Mission at Nursia, in Australia, I conceived, that some mistake had been made in 1845, by the missionary who sent in the report.[5] But later, when I began to notice the coincidence of testimony from many quarters, in many ages, then I could not conceal from myself that this chapter must be read. It is in conflict with our prevalent theories of the development of gods out of worshipped ancestral spirits: for the maker of things, not approached in prayer as a rule, is said to exist where ancestral spirits are not reported to be worshipped. But science (in other fields) specially studies exceptional cases, and contradictory instances, and all that seems out of accord with her theory. In this case science has glanced at what goes contrary to her theory, and has explained it by bias in the reporters, by error in the reporters, and by the theory of borrowing. But such coincidence in misreporting is a dangerous thing for anthropology to admit, as it damages her evidence in general. Again, the theory of borrowing seems to be contradicted by the early dates of many reports, made prior to the arrival of missionaries, and by the secrecy in which the beliefs are often veiled by the savages; as also by the absence of prayer to the most potent being.