no more than a suggestive and ingenious guess. As such, however, it is certainly one to be borne in mind in future investigations.
The argument from customs is the weakest. Twenty-one customs are enumerated on p. 27 as common to the Tasmanians and Australians; and the author says: "This list of remarkable practices, identical in both countries, is surely sufficiently imposing to establish of itself a very intimate connection, if remote in time." If these twenty-one customs were peculiar to Australia and Tasmania his argument would be valid. This he does not seek to prove; and if he did he would fail, for thirteen at least are common to savages over the greater part of the world, and four or five others are known in other lands than Australia and Tasmania, leaving only three or four at the most—more probably two or even one—possessing any real significance. Chapter VIII. contains an interesting discussion on the question whether group-marriage has ever prevailed in Australia. The author sets himself in opposition on this point to the opinion of Messrs. Fison and Howitt, whose arguments, reinforced by Messrs. Spencer and Gillen, seem to me to be conclusive. This chapter also contains some useful tables of the marriage systems of various tribes, and some acute observations which deserve to be considered; but I hardly think he has fathomed the difficult subject of the marriage regulations. The remaining chapters, IX. and XL, in which the argument from customs is elaborated, are thin. They are vague and add nothing to our knowledge. It is obvious that Mr. Mathew has not made a study of savage belief and practice. Chapter X. reproduces a number of drawings found upon rocks and in caves in various parts of Australia. Most of them, if not all, have been published before. Some of them have been a puzzle to anthropologists ever since they were discovered, as being above the level of art of which the aborigines at present seem capable. It cannot be said, however, that Mr. Mathew has thrown much light upon them. His reading of what look like alphabetical characters is very doubtful; and his citations from Moor's Hindu Pantheon cannot be taken seriously.
