adduced, from China. Still, these examples are a feeble support for the weight of the author's conclusion.
But it is when Dr. Trumbull comes to the origin of the threshold rite that we part company with him. To find the origin in the act of nuptial union is nothing but a guess. It is true that scientific men have occasionally arrived at the true solution of a problem by a bold guess. The instances are rare; and guessing is not to be commended as a method of attaining truth. Dr. Trumbull disguises his guess as an "induction." But the induction starts from a series of assumptions as to "man's earliest religious ideas" and his earliest social relations, which do not conform to the results of scientific investigation. It may be that these assumptions were necessary, or believed to be necessary, to a certain theological position. We doubt that they are really so. At one time it was thought necessary to believe that the world was flat, or that it was created in six days of twenty-four hours. Nobody now holds these doctrines; but the theological position is practically unchanged. So by-and-by theologians will recognise that the anthropological view of human origins and human civilisation is the correct view. Dr. Trumbull cites Professor Robertson Smith and Mr. Frazer. If he had followed these masters, and others on his own side of the Atlantic, without flinching, he might have made an important discovery; and we believe he would ultimately have carried his professional brethren with him. We regret that he has not done this. He has written a very interesting book; he has abundantly demonstrated the sanctity of the threshold,[1] but he leaves the reason for that sanctity as dark as he found it.
It may be an insular prejudice, but we cannot approve of the supplement consisting of comments of specialists to whom the proof-sheets were submitted. It is unfair alike to the specialists themselves and to the readers. Dr. Trumbull would have done better to consider the comments in the privacy of his own study, and to revise his work in accordance with the more valuable of the suggestions thus obtained.
