antiquity and with the traditions of other quarters of the globe is that they frequently suggest to the student subjects for further inquiry.
The same inequality is even more observable in his arguments and analyses. He often sums up the facts with precision and states the result correctly and with clearness. Nothing could be more admirable, for example, than his demonstration of the historical impossibility of the legend of St. Martha and the Tarasque, or his exposition of several of the credulities and practices with which he deals. Had he confined himself to a selection of these and wrought them out in detail his book would have been more useful. But the truth is he covers too large a ground; he ventures upon the discussion of customs and myths to which he has paid too little attention; and he is unacquainted with the results of recent investigation. Did space permit, many illustrations of these shortcomings might be given; it must suffice to refer to the chapter on Marriage and Progeniture. Here he relates a number of practices in Provence, some of which appear to be relics of Priapean worship, others of quite different rites for procuring children, others of such marriage-customs as bride-capture, bride-recognition, and so forth, others again of savage ante-nuptial licence. All these he mixes up together as vestiges of primitive hetairism, and fortifies them with references to other practices, relevant and irrelevant. The careful examinations of marriage customs undertaken by Morgan, M'Lennan, Westermarck, and others, it seems, have hardly yet penetrated into the scientific circles of Toulon. Dr. Bérenger-Féraud at least gives no sign of having heard the names of any of these writers, though he is acquainted with Sir John Lubbock's Origin of Civilization (a work, important as it is, yet constituting by itself an insufficient armoury for one who sallies forth to attack so hard a problem), and he mentions Letourneau in another connection.
A characteristic which we are a little surprised to find in a writer who has watched the savage in his own wilds is the disposition to ascribe far too large an influence to deliberate fraud. Fraud is accountable for something in the religious history of mankind. But even fraud must have a basis of pre-existing belief to build on: it presupposes the superstition of which it takes advantage; and the wider the knowledge of savage belief the smaller is the part one is inclined to admit for fraud, and the larger the part
