ON ETHNOLOGY. 
327 
of Sanscrit grammar. Schleiermacher for instance, in his ‘ Lesay on the In¬ 
fluence of Writing on the Language,’ says:—“Or cest une question de 
savoir si jamais lo peuple du Bengale, desceiidu probablement de ces habi- 
tans primitifs, s’appropriat eiitidretnent le Sanscrit, ou bien s il ne se format 
pasdeslecommenccnKUt deson a^RujctllMment ades 6trangors plus civilis6s 
quo lui, une langue plus scniblablc a sou idionie actuel quau Sanscrit, en 
suppriinant les tbmics graramaticales trup difficilcs dc celui-ci ct n’adopt^t 
que les mots. Ou ue peut pas pr£*teudre qull est invraisemblable, qu’un 
peuple tout entier ait accept^ ridimiic de quelques prctres et guerriers car 
nous avoQS vu cela s’ettectuer au royamne d’Awuirn, qui dans les temps 
aociens doit avoir fait partie des pay* Himlous. Cependant plus tard on y a 
eu jusqu’au commencement du dix-septieme si^ele uue langue, des livres et 
des institutions scmblables « crllt-a des pays au-dclu du (iauge; mais alors 
les Brahmans s’introduisirent dans cc pays, y rCqaiiidircnt leurs doctriues, 
convertirent le roi, ct bieutot nn dialecte Bengale rcMipla^a I’ancien idiome, 
de sortc que celui-ci appartient mainlcuant presque entidrement aux langues 
raortes.” 
Now one must admit, tliat if such a fact could be proved, that the gram¬ 
matical elements of the Bengali arc not originally Sanscritic, but belong to 
another system of languages, this would change entirely the view which we 
have taken of the ethnographic and linguistic relations of the inhabitants 
of India, and bring on the same confusion as if it could be proved that the 
grammatical system of the oioderu Persian M'as not of an Iranian character, 
or that the F.uglish language was not of a Teutonic origin. For why do we 
call the English language a Toutouic one, if not because the Saxons, settling 
in Britain, did not change the grammatical character of their language, al¬ 
though they adopted many wonls from the Celtic nation which they sub¬ 
dued ? And wliy do we not say, that after the Norman conquest the 
language of Enclaml became a Norman language, if not because, though 
overgrown with Norman words, it preserved its own grammatical system ? 
Why ai’o the Romance called Romance, and the Teutonic, Teutonic lan¬ 
guages ? Because the I’eutonic race, when brought into contact with 
Roman civilization and language, found iu its strongly developed nationality 
sufficient strength lo appropriate arul incorporate into its language a great 
number of Latin words, without giving up the essentially Teutonic form of 
iw grammar, while the Celtic nations yielded to the overpowering influence of 
the Roman civilization, and adopted not only tlie substantial but also the 
formative element of the Latin language, thus giving rise to new languages, 
which cannot be considered as Celtic idioms, but as branches of the Latin 
language, modifieti and developed by Celtic elements. 
Looking then from this point of view at the question about the origin of 
tlie gratnmatical forms in the ntodem Indian languages, I tliought it neces¬ 
sary to take this subject into serious consuleration. I was convinced that it 
would be possible, either to account for the heterogeneous influence which 
ha-s been acting upon the Imiguages of the ludian nations, or to find a con¬ 
necting link betwi^en the grammar of the old and modern Indian dialects. 
The results of my inquiry I have laid down in a Comparative Grammar of 
the Bengali Language, and I shall quote thence a few points in answer to 
a theory which has been proposi^d in regard to the grammatical structure of 
the Bengali language by the Rav. Dr. Stevenson, whose extensive knowledge 
of dialects spoken in India, which be hail tlie opportunity of studying on the 
spot, entitles his riews to great attcutiou and careful examiuatioQ. 
1 quite agree with Dr. Stevenson in the manner in which he tries to 
prove that there existed in India an aboriginal language dififerent from the 
