The hymns of the Rig Veda inextricably confused; the deities of an earlier era confounded, and again merged together in a pantheism now complete; the introduction of strange gods; recognition of a hell of torture; instead of many divinities the One that represents all the gods, and nature as well; incantations for evil purposes and charms for a worthy purpose; formulae of malediction to be directed against 'those whom I hate and who hate me'; magical verses to obtain children, to prolong life, to dispel 'evil magic,' to guard against poison and other ills; the paralyzing extreme of ritualistic reverence indicated by the exaltation to godhead of the 'remnant' of sacrifice; hymns to snakes, to diseases, to sleep, time, and the stars; curses on the 'priest-plaguer'--such, in general outline, is the impression produced by a perusal of the Atharvan after that of the Rig Veda. How much of this is new? The Rig Veda is not lacking in incantations, in witchcraft practices, in hymns to inanimate things, in indications of pantheism. But the general impression is produced, both by the tone of such hymns as these and by their place in the collection, that they are an addition to the original work. On the other hand, in reading the Atharvan hymns the collective impression is decidedly this, that what to the Rig is adventitious is essential to the Atharvan. It has often been pointed out, however, that not only the practices involved, but the hymns themselves, in the Atharvan, may have existed long before they were collected, and that, while the Atharvan collection, as a whole, takes historical place after the Rig Veda, there yet may be comprised in the former much which is as old as any part of the latter work. It is also customary to assume that such hymns as betoken a lower worship (incantations, magical formulae, etc.) were omitted purposely from the Rig Veda to be collected in the Atharvan. That which eventually can neither be proved nor disproved is, perhaps, best left undiscussed, and it is vain to seek scientific proof where only historic probabilities are obtainable. Yet, if a closer approach to truth be attractive, even a greater probability will be a gain, and it becomes worth while to consider the problem a little with only this hope in view.