240 
REPORT—IS47. 
same stock. The following observations will perhaps explain as briefija 
possible the principles wiiich have either been expressed or followed taedf 
by pliilologists who have entered upon such inquiries. 
It is the prevalent opinion of philologists that the most extensive relalioii 
between languages and those which are tlic least liable to be effaced by tia 
and foreign intercourse, are the fundamental laws of construction bothlj 
words and sentences. Grammatical construction, or the rules whidi 
the relation of words in sctitcnces, appeal's to be very enduring and coosuii 
since a similar construction prevails through whole classes of langiups 
whicli have few words in common, though they appear originally to hiw 
had mure. But beyond this there is a cognate character in words dit^ 
selves, which sometimes pervades the entire vocabulary of a whole family 
languages, the words being ftrrmed in the same manner and accordiDi '• 
till! same iirtifichil rule. 'This may be exemplified in the monosTllil^ 
structure of the Chinese and Indo-C’hinefie languages, and by the 
of the vocalic harmony pervading the languages of High-Asia. wiiict l 
shall presently describe, and perhaps by the dibsyllnbic structure ot rMW® 
the Syro- Arabian languages. Of grammatical analogy or similarity in 
laws ot construction of words in sentences, including the rules ot 
we liave examples in the Itmguagos of the ulmriginal American iiations.M| 
perhaps tliu most remarkable sjiccimcn is to be found in the graniiuai'^^ 
system of the Indo-European languages. 
It has been doubted, and the qm^stion is difficult to decide, 
analogy of structure alone is sufficient to prove emnnmnity of origin w* 
inuorted bv verbal rescuiblain'^' 
tween languages,—1 mean when it is unsupported by v 
Languages descended from the 
_ . same stoeic as it would seem, 
some words in common, or at all ('vents some similar words. Sucliw^ 
may be few in number, as wc know that they arc iu idioms 
longing to one family. If we compare the Welsh and the Kussiaa, ^ 
Indo-LurojH:an laugnages, we shall linil few similar words. On the 0 
hand, the existence of a considerable number of similar words in anyj” 
languages docs not prove those languages to be of cognate origin* 
many words has the Welsh borrowed from the English and the t'*?. , 
from the I-rencli! These idioms arc indeed descendants of one oi'P 
stock, but their relation is distant and has no connection with tbe 
tioiis which they have respectively made, one from the vocabulaiy 0 ' _ 
other, in recent times. The evidence to be deduced from verbal an*^ 
depends much on the classes of words in which such analogy is to bo 
and on this point the following observations liave been made*. 
^ Ine yiarticular classes of words whicli resemble each other in 
roii-iouiar Classes ol words whicli resemble each other m 
of a coimnon origin and derived from one primiUve stock, are 
erent in kind from those u’hicU one nation borrows from its neig 
Even wiion one. people has derived from another a considerable 
ot Its entire stock of words, there generally remains an indigenous ora^ 
gina vocabulary, or if I may be allowed the expression, ahome-bred _ 
consisting of such words as children icani in curly infancy, and in 
veloperaent of thoir faculties. This domestic vocahularv consists of 
« I-I such OS those denoting family relations, ‘father, 
.p", ’ brother, ‘ sister;’secondly, words denoting >-arious parts of tb® • 
thirdly, names of material and visible objects and the elements of 
the heavenly bodies, Ac. ; fourthly, names of domestic animals; JT 
‘Sl I'odily acts, such as ‘eat,’ ‘drink,' ‘sleep. 
’ *^*; sixthly, personal pronouns, which are found to be aaioflg 
* Abel Remusat, Recherches sur les Langues Tartares. 
