472 COSMOS. 



most nearly the language common to all who had migrated 

 from the general seat of origin. The largest field for such 

 investigations into the ancient condition of language, and 

 consequently into the period when the whole family of man- 

 kind was, in the strict sense of the word, to be regarded as 

 one living whole, presents itself in the long chain of Indo- 

 Germanic languages, extending from the Ganges to the 

 Iberian extremity of Europe, and from Sicily to the North 

 Cape. The same comparative study of languages leads us 

 also to the native country of certain products, which, from 

 the earliest ages, have constituted important objects of trade 

 and barter. The Sanscrit names of genuine Indian products, 

 as those of rice, cotton, spikenard, and sugar, have, as we 

 find, passed into the language of the Greeks, and, to a certain 

 extent, even into those of Semitic origin.* 4 



From the above considerations, and the examples by which 

 they have been illustrated, the comparative study of languages 

 appears as an important rational means of assistance, by 

 which scientific and genuinely philological investigations may 

 lead to a generalisation of views regarding the affinity of 

 races, and their conjectural extension in various directions 

 from one common point of radiation. The rational aids towards 



* In Sanscrit, rice is vrilii, cotton karpasa, sugar 'sarkara, and 

 spikenard nanartha; see Lassen, Indisclie Altertlmmskunde, bd. i. 

 1843, s. 245, 250, 270, 289,<and 538. On 'sarkara and kanda (whence 

 our sugar-candy), consult my Prolegomena de distributione geogra- 

 phica Plantarum, 1817, p. 211. " Confudisse videntur veteres saccha- 

 rum verum cum Tebaschiro Bambusse, turn quia utraque in arundinibug 

 inveniuntur, turn etiam quia vox sanscradana scharkara, quse hodie (ut 

 pers. schakar et hindost. schukur) pro saccharo nostro ; adhibetur, obser- 

 vante Boppio, ex auctoritate Amarasinhae, proprie nil dulce (madu} 

 significat, sed quicquid lapidosum et areuaceum est, ac vel calculum 

 vesicse. Verisimile igitur, vocem scharkara initio dumtaxat tebaschi- 

 rum (saccar nombu) indicasse, posterius in saccharum nostrum humili- 

 oris arundinis (ikschu, kandeksclm, kanda), ex similitudine aspectua 

 translatam esse. Vox Bambusae ex mambu derivatur ; ex kanda nostra- 

 tium voces candis zuckerkand. In tebaschiro agnoscitur Persarum 

 schir, h. e. lac, sanscr. kschiram" The Sanscrit name for tdbaschir is 

 tvakkschira, bark-milk; milk from the bark. See Lassen, bd. i. 

 s. 271-274 ; compare also Pott, Kurdische Studien in the Zeitsdirift 

 fur die Kunde des Morgenlandes, bd. vii. s. 163-166, and the masterly 

 treatise by Carl Bitter, in his Erdkunde von Asien, bd. vi. 2, s. 

 232-237. 



