94 Reviews.
proof, the authors find the way of perfection in the study of typical facts. These they borrow mainly from the Sanskrit texts and from the Bible. Nowhere else are recorded the exact steps of the ritual with sufficient accuracy and sequence. To look for their typical facts in ethnological collections would, they hold, be to carry their study over groups of facts artificially formed, and not in their ordinary and actual growth. In the definite and complete rituals with which they deal, they have totalities of natural systems of rites ; and by keeping within the restraints imposed by the texts they are the less exposed to omissions and to arbitrary classifications. But they do not refuse to appeal either to classical sources or to ethnological collections to illustrate their analyses and to check the generality of their conclusions. Lastly, as the two religions with which they chiefly concern themselves are very different, the one tending to monotheism* and the other to pantheism, they hope in comparing them to arrive at suffi- ciently general conclusions.
Here I must interpose an observation or two. No doubt such an examination as the authors have carried out will yield, and in fact does yield, a large amount of valuable information about the objects and method of sacrifice in its most developed forms. In- cidentally this may afford us grounds for various conjectures about the origin and early forms of sacrifice ; and after all it is the origin and early form.s which are the most important in this inquiry. It may be that nothing more certain can be obtained than the con- jectures we may make from observation of these finished forms. But we cannot assume that, because Robertson Smith's method may have been arbitrary and Tylor may have neglected an essential aspect of the problem, an examination of the forms of sacrifice current in other countries and other grades of civilisation (though our information about them may be less full, may even be often defective in serious particulars), will not result in the discovery of facts pointing to a theory yet nearer the truth. I for one cannot admit that an occasional appeal to ethnological sources to control the conclusions arrived at by the examination of Hindu and Hebrew sacrifices is at all sufficient, or is less arbitrary than the method to which the authors rightly or wrongly object. The most accurate analysis of these sacrifices will not absolve the student who desires to attain a reasonable and fairly verifiable theory on the origin and early forms of sacrifice from examining in detail the procedure
