archæology. Savage life, carrying on into our own day the life of the Stone Age, may be legitimately claimed as representing remotely ancient conditions of mankind, intellectual and moral as well as material. If so, a low but progressive state of animistic religion occupies a like ground in savage and in primitive culture.
Lastly, a few words of explanation may be offered as to the topics which this survey has included and excluded. To those who have been accustomed to find theological subjects dealt with on a dogmatic, emotional, and ethical, rather than an ethnographic scheme, the present investigation may seem misleading, because one-sided. This one-sided treatment, however, has been adopted with full consideration. Thus, though the doctrines here examined bear not only on the development but the actual truth of religious systems, I have felt neither able nor willing to enter into this great argument fully and satisfactorily, while experience has shown that to dispose of such questions by an occasional dictatorial phrase is one of the most serious of errors. The scientific value of descriptions of savage and barbarous religions, drawn up by travellers and especially by missionaries, is often lowered by their controversial tone, and by the affectation of infallibility with which their relation to the absolutely true is settled. There is something pathetic in the simplicity with which a narrow student will judge the doctrines of a foreign religion by their antagonism or conformity to his own orthodoxy, on points where utter difference of opinion exists among the most learned and enlightened scholars. The systematization of the lower religions, the reduction of their multifarious details to the few and simple ideas of primitive philosophy which form the common groundwork of them all, appeared to me an urgently needed contribution to the science of religion. This work I have carried out to the utmost of my power, and I can now only leave the result in the hands of other students, whose province it is to deal with such evidence in wider schemes of argument. Again, the intellectual rather than the emo-
