early, and many women come into the genealogies, who as a rule, marry very early. It is this latter number, therefore, that will be adopted in fixing dates in what follows.
As a rule the Polynesian genealogies are reliable within certain limits and go very far back. I cannot at all agree with Mr. Basil Thompson[1] that they "do not carry us back for more than seven or eight generations, and beyond this limit we are apt to step into the regions of mythology." This is a very surprising statement to emanate from one who has passed some years amongst various branches of the race, i.e., Tongans, and the half-caste Polynesians of Fiji. To those who have studied this question amongst various branches of the race, no proof is necessary as to the general accuracy—always within certain limits—of Polynesian genealogies; but as Mr. Thompson has—very rightly—acquired a good deal of fame by his writings, it is necessary to show that his estimate of the value of these genealogies is mistaken. At the same time, for the purpose I have in hand, some evidence is also requisite, in order to judge of the degree of reliability that may be placed on the dates to be used herein. First may be mentioned, that the great migration to New Zealand took place at 21–22 generations back from the year 1900, or in the year 1350. This date is arrived at by taking the mean number contained in over 50 genealogical tables going back to those who came here in the fleet, all of which will agree to within 4 or 5 generations in number. Where many women come into the lines, they are naturally longer.
But the severest proofs that can be applied to these tables, is to compare those from different branches of the race showing descent from the same ancestor. The first
