dilapidated tabernacle. This explanation of the rule that the King of the Wood had to be slain by his successor at least renders that rule perfectly intelligible. Moreover it is countenanced by the analogy of the Chitombé, upon whose life the existence of the world was supposed to hang, and who was therefore slain by his successor as soon as he showed signs of breaking up. Again, the terms on which in later times the King of Calicut held office are identical with those attached to the office of King of the Wood, except that whereas the former might be assailed by a candidate at any time, the King of Calicut might only be attacked once every twelve years. But as the leave granted to the King of Calicut to reign so long as he could defend himself against all comers was a mitigation of the old rule which set a fixed term to his life, so we may conjecture that the similar permission granted to the King of the Wood was a mitigation of an older custom of putting him to death at the end of a set period. In both cases the new rule gave to the god-man at least a chance for his life, which under the old rule was denied him; and people probably reconciled themselves to the change by reflecting that so long as the god-man could maintain himself by the sword against all assaults, there was no reason to apprehend that the fatal decay had set in.
The conjecture that the King of the Wood was formerly put to death at the expiry of a set term, without being allowed a chance for his life, will be confirmed if evidence can be adduced of a custom of periodically killing his counterparts, the human representatives of the tree-spirit, in Northern Europe. Now in point of fact such a custom has left unmis-
