THR ETHNOLOGY oF THK INDIAN aUCHIPKLAOO li 



ha* recently returned to fhe subject with ft great accession of 

 important facts derived fmm a laborious comparison of entire 

 (Jfrtrnnarica of the Javanese Malay, Bugi*, Tagnla, Bisavu 

 Tal. »iiun, New Zealand and Mnla-jasT. * His opinion remains un- 

 uftectctl by the n*earchea of Humboldt. 



The historv, of ihe Polynesian trihes hu* frequently been made 

 the fnbject of vepai-aie enquiry and speculation. Different writer 

 derive tbero from ^America, the northern part of the Pacific, ami 

 the Indian ArchijAkgO. A mongM recenf enquirers Mr Will tarn* 

 has strongly and ably maintained their Indonesianf orierin, and Mi 

 Half, to whose important and valuable labours 1 ehnfi afterward* 

 fully advert, has adopted the same views. Following up Mr Wil- 

 liams' connection of the Society with the Sandwich IdlanAs, through 

 the identity in ihe name of two fehuidt, Hawuii, he has »hewn that 

 this name may be traced throughout oil the greater grottos, 

 applied to important islands m which the race m !o ;ed, to a tradi- 

 tional Jnnd of origin, or to a lower region the abode of departed 

 spiriLn. Mr Wiltlnm? in his work had interred thnt the Societv wen- 

 peopled from (he Sandwich Wand*, but whi n Mr fr.de mentioned 

 Ut him the probable use that mipht be made of this tmrnt in referring 

 baek the different tribes to their original 9enf, be told him 'Jiat he had 

 long entertained the opinion, that theSamoan islands were the tourv- 

 of the population of the other groups of Polynesia. Mr William t- 

 was not awttj-c that the mime of the larj/ext island of the latter 

 group is Savaii, which Mr Hale concludes to have been i\» 

 original of all the other Savaib throughout Polynesia. : In hi* 

 chart of Oceanic migrations he does not allow any line of connec- 

 tion from Micronesia to Polynesia, but abruptly "mite off on tlV 

 N.B. verge of Melanesia the stream that, according to him, 

 people*) the western groups of Polynesia and thence flowed 

 to the B*,El, and N. clusters. He conjectures however that ih. 



! i 1 , 011 M * ia ) an i*olja«stan Languages and Race* ■ /our. tad. Arch 

 vo!. II. p. 1BH. 



t The name Indian Archipelago is too long to admit of being oned in an adleetf*. 

 or m ut etlwetfraphical form. Mr liart soffgusts ths Uhnouraphicai term Intta 

 i^atana hut ruecti if in favour of Maiaynn^ans, (ante p. 71). Fur reason* 

 -rfclrh wilt be obtfous on reading a •n&n*,utm note, f prefer th* purrty geographical 

 t'Tto Indent-sis. whkh ia inercfy a shorter synonym far the Indum island* or 

 rke Indian Aichiptbgo. We that get Induneaian for Indian ArchLpdagiau or 

 uchipdagie, ami tadon«sianj fat radian Arrhi[*lsgian» »r Intiimi iilandi-r* 

 i have no afTerffon Sr the mnJUpHeutJen ofwrml-grrelan wordn, and would gladl) 

 are all the mmaf wiped olf the mopif good 8«on *-qul*al«its could he iubstituted. 

 I iu itrm Jin* some claim however to he located in die region, lor In the sHgbtJv 

 d'ttrenL form oi nasn It is perhaps as ancient in die Indian Arthipeligo bi in 



* I ivee«». 



t The regular phonetic changes which th t word naturally undergoes In the diftr- 



-ill dUlccy axe ' 



Saiuoan dialect - Sftvoii 



Tahitian - HavoiL 



SaD.Jwj.ii I. _ i, ■ /-.;■ 



Uaratongan - Avaiki. 



Nukuhiran MavnikL 



New Zealand tfctwriAi. 



i 



