THE PAUMOTU IN THE POLYNESIAN SCHEME. 55 



Paumotu-Tahiti : 



961 1029 1130 1203 1246 1310 1360 141 1 1465 1537 1572 161 1 1651 1682 



966 1033 1132 1204 1252 1311 1361 1412 1478 1538 1574 1612 1652 1683 



967 1038 1139 1208 1255 1313 1364 1416 1490 1539 1576 1613 1653 1686 

 969 1046 1148 1211 1257 1314 1365 1418 1495 1540 1577 1614 1654 1688 

 973 1055 1149 1214 1258 1318 1369 1419 1496 1542 1578 1615 1658 1690 



976 1062 1151 1215 1259 1325 1370 1420 1498 1543 1579 1616 1659 1691 



977 1064 1155 1216 1262 1326 1376 1421 1509 1544 1580 1618 1660 1698 



978 1065 1165 1219 1263 1327 1379 1422 1510 1545 1581 1619 1661 1699 



979 1068 1 166 1220 1269 1330 1380 1423 1515 1549 1585 1623 1663 1703 



981 1069 1171 1221 1272 1331 1383 1424 1516 1552 1586 1627 1665 1711 



982 1073 1174 1222 1273 1340 1384 1425 1517 1553 1587 1628 1666 1714 

 986 1077 1175 1223 1275 1342 1390 1426 1519 1554 1591 1629 1668 1721 

 993 1079 1176 1226 1281 1345 1393 1427 1520 1556 1592 1630 1669 1722 



998 1085 1 180 1230 1283 1346 1396 1429 1521 1559 1595 1631 1671 1729 



999 1093 1 181 1232 1290 1351 1398 1437 1524 1562 1596 1638 1672 1730 

 1000 1 100 1 183 1235 1293 1353 1402 1438 1525 1563 1598 1639 1674 1731 

 1008 1 1 10 1 188 1236 1299 1354 1405 1449 1526 1564 1599 1640 1675 1732 

 1010 1 1 17 1 191 1242 1300 1355 1407 1450 1533 1565 1602 1641 1676 1734 

 1012 1127 1198 1243 1305 1356 1409 1451 1534 1566 1603 1643 1679 1735 

 1015 1 128 1200 1244 1308 1357 1410 1463 1536 1567 1604 1646 1680 



Finally a few brief tables will disclose the tale that this arid numerical 

 waste has to tell. 



The material available for the foregoing study of the Paumotu is 

 summed in 2,550 items. Of these we have developed identifications in 

 other Polynesian tongues for 1 ,335 items, 52 per cent. Of this Paumotu 

 element 577 items reveal their affiliations in this province of Southeast 

 Polynesia, 43 per cent of Paumotu speech. Similarly we find 758 items 

 whose affiliates are in the Polynesian of the archipelagoes westward 

 and earlier along the migration track, 57 per cent. In the more minute 

 study of affiliation we see that 455 items are identifiable in Rapanui, 

 34 per cent; 1095 in Tahiti, 81 per cent; 583 in Mangareva, 42 per cent; 

 645 in the Marquesas, 48 per cent. 



In the preceding paragraph I have first established the percentage 

 of affiliates in bulk. Thereafter I have established the percentages 

 through the use of 1,335, the sum of the identifications, as the denomi- 

 nator. Of course it is possible for those students who prefer it to estab- 

 lish the percentages in bulk by the employment of denominator 2,550; 

 the relative proportion will not thereby be affected, for the ratio, once 

 established, is constant. 



In defense of my method I suggest the following considerations. 

 Through initial dichotomy we have established two classes in the Pau- 

 motu: that in which exterior affiliation is discovered, that in which 

 such affiliation has not yet been discovered two classes nearly equal 

 in extent. We must consider the position of the unidentified class. At 

 present it stands simply as speech material peculiar to the Paumotu. 

 We then meet the problem, is this peculiar possession Polynesian or 

 alien contamination? If alien contamination, whence comes it? 



There is not an item in this class which might not be Polynesian, 

 firmly established by its occurrence in no more than a single outer lan- 

 guage of the family. Form, usage, sense-structure, all conform rigidly 



