one feels constrained to assert for pure good faith and simple folly, in the development of human superstition.
It irks us to write thus of the book before us, for we have found much in it alike of interest and instruction. We have abundant reason, moreover, to be grateful to Dr. Bérenger-Féraud for earlier works deservedly prized by scientific students. We only regret that he has marred a well-intentioned effort by the ambition to do too much. After all, this is a vice that leans to virtue's side.
The correction of the proof-sheets has been far from exhaustive; and the spelling of foreign words has proved too great a task for the printer's devil. We hope that philological speculations will be excluded from the remaining volumes. They are always dangerous; and where even the greatest authorities sometimes blunder an amateur can only venture at the risk of serious and repeated mishaps.
The first series of these legends was reviewed in last year's Folk-Lore (p. 391). The second is like to it, though it contains fewer "origins" and more witch-stories. On this occasion the ingenious author has complied with the request of his reviewers, to declare which parts are literally translated from his authorities; but this not without a gibe or two. In one or two stories he has given a new form to the incidents, although he has not altered these (pp. 121, 204); but for the most part the tales and verses are given as he received them.
A large number of incantations are given, and some of the tales are obviously composed "by witches for witches," merely to explain and preserve these. One charm tells how to become male or female at will (p. 49).
There is a vast deal about witches and witchcraft; when present at a baptism they appropriate part of the blessing (p. 28); after death they go not to hell, not to heaven, not to purgatory (p. 101); we learn how they "bind" a bridegroom, and how he is cured (p. 202). Diana appears as queen of the witches, in love with her brother (p. 209).
As regards origins, the most important legend, and a very
