This, briefly stated, is the hypothesis of De Brosses.  If he had possessed our wider information, he would have known that, among savage races, the worships of the stars, of the dead, and of plants and animals, are interlaced by the strange metaphysical processes of wild men.  He would, perhaps, have kept the supernatural element in magical stones, feathers, shells, and so on, apart from the triple thread of Sabaeism, ghost-worship, and totemism, with its later development into the regular worship of plants and animals.  It must be recognised, however, that De Brosses was perfectly well aware of the confused and manifold character of early religion.  He had a clear view of the truth that what the religious instinct has once grasped, it does not, as a rule, abandon, but subordinates or disguises, when it reaches higher ideas.  And he avers, again and again, that men laid hold of the coarser and more material objects of worship, while they themselves were coarse and dull, and that, as civilisation advanced, they, as a rule, subordinated and disguised the ruder factors in their system.  Here it is that Mr. Max Muller differs from De Brosses.  He holds that the adoration of stones, feathers, shells, and (as I understand him) the worship of animals are, even among the races of Africa, a corruption of an earlier and purer religion, a ‘parasitical development’ of religion.

However, Mr. Max Muller himself held ‘for a long time’ what he calls ’De Brosses’s theory of fetichism.’  What made him throw the theory overboard?  It was ’the fact that, while in the earliest accessible documents of religious thought we look in vain for any very clear traces of fetichism, they become more and more frequent everywhere in the later stages of religious development, and are certainly more visible in the later corruptions of the Indian religion, beginning with the Atharvana, than in the earliest hymns of the Rig Veda.’  Now, by the earliest accessible documents of religious thought, Professor Max Muller means the hymns of the Rig Veda.  These hymns are composed in the most elaborate metre, by sages of old repute, who, I presume, occupied a position not unlike that of the singers and seers of Israel.  They lived in an age of tolerably advanced cultivation.  They had wide geographical knowledge.  They had settled government.  They dwelt in States.  They had wealth of gold, of grain, and of domesticated animals.  Among the metals, they were acquainted with that which, in most countries, has been the latest worked—­they used iron poles in their chariots.  How then can the hymns of the most enlightened singers of a race thus far developed be called ‘the earliest religious documents’?  Oldest they may be, the oldest that are accessible, but that is a very different thing.  How can we possibly argue that what is absent in these hymns, is absent because it had not yet come into existence?  Is it not the very office of pii vates et Phoebo digna locuti to purify religion,