142 Ethnological Data in Folklore,
important question to ask : Are there customs which will not produce the test-figure? For the purpose of answering this I have compared roughly the important group of cus- toms relating to water-worship.
Now I have stated in my Ethnology in Folklore rea- sons for considering water-worship customs to be non- Aryan in origin, to belong therefore to the pre-Celtic people of these islands ; and it is remarkable that the " geographical test-figure " produced from the water-cus- toms differs radically from that produced by the fire-cus- toms. I suggest therefore that in this interesting fact we have provisionally a proof of the value of this method of studying the ethnological basis of folklore.
But there is another great peculiarity which distinguishes the group of water superstitions from fire superstitions. Water superstitions never lead up to a tribal organisation, nay, they oppose tribal organisation. They correlate a set of ideas inconsistent with tribal organisation. The conclusion therefore is irresistible that water superstitions do not belong to the Aryan Celtic people, who were tribal, or the Aryan Teutonic people, who were tribal ; from which it follows that they must belong to a pre-Celtic non-Aryan people.
I attribute very great importance to the interposition of the tribe at some point in the history of a given item of folklore. It is vital ; not accidental. Let me for one moment turn to marriage customs in British folklore. I have studied these for some years, and hope to be able to lay before the Society a few results at no distant day. They divide sharply off into two distinct groups. One group leads us to the tribal organisation ; can be explained only by the tribal organisation of Celts and Teutons. The other group leads us to a social organisation, which is I would almost say anti-tribal. Of course, it may be argued that this second group consists of the worst worn fragments of the former group. But against this theory is the fact that they are of themselves a determinative group, and not
