142 RTlINOLOOr OP THK IN DO-PA CTFIC ISLAKBS. 



quite differenfc, although they also Imve Scytliic affinities. It eeeiiis 

 probable that the race which gavo tliD horse to the ancient Indians 

 or to India was at one period an isiflueDtial one in Middle Asia, 

 and that its movements extended to the Caucasian province on 

 the weat aud to the Indus on the east. 



The Dmvirians could not have carried the horse t-o Asoncaia in 

 the Indo-Australiaii era, and it was probably not till long after ihcy 

 possDsaed it that their inaritimo art became sufficiently advauccd 

 to enable tbem to transport it to the islands. It is remai'kable tbat 

 no trace of the South Braviriao luiine esists iu Ultraindia, al- 

 though it is found in Sumatra, kuda, a name which has been car- 

 ried by the Malays over a large portion of Indonesia. The proba- 

 bility ia that this tenn waa iiitrodiiced directly from Southern India 

 by the Kalingaa after the ship-building period of the Drai'irian 

 civilisation commenced, and that it is of a similar age to the 

 ]VIalay kapal, "ship," The Hindi ghora ia now the moat prevalent 

 term not only in the Himalayan bxit in the Ultraindian languages. 

 Prior to its introduction, however, the Gangetic lang\mge3 had 

 anotber torm for the horse and it appears to have been carried to 

 Indonesia before the South Dravirian kuda, 



2. payiffw, Tamil ane. This exceptional term appears to be an 

 application of a native term for "cow** to the horao ( see Cow 1, ' 

 Mai. paya, Karp;i payii). 



8, saflliam, sadam Eol, ( P Gond chuddur ), Kewar sala> Chepang 

 aerang, Milchanang rang, Lungkho rang, rung, Burm, j/trang, 

 wi ven, Siugplui Ara-?»rang. This term appeara to liave preceded the 

 Braviriau kuda in Indonesia. It is found further caatward than 

 Sumatra in the vocahulariea of all the civilised tribes. It appears 

 to have been carried from Java to Celebes, and from both aa cen- 

 tres to other islands. The prevalent forms are jaran, jara, cbaran, 

 ajaraug, jaraiig, ajura, anyarang, nyarang, adala, ndala. 



i^rom tho distribution of these forms there can be little doubt 

 tbat the Forth Gangetic saraug, aala, rang^ ia the original and that 

 rang or ra, la, is the ultimate root. It appears to have been the term 

 in use by the most civilised nation of the Granges at one period, 

 to havo been communicated by it to the hill tribes on botli sides, 

 aud to have bocu carried in the cfjurse of its commerce to iriLra- 

 india and Java. The Indonesian forms, it will be remarked, are 



