ON ETHNOLOGY. 
313 
nevertheless is preserved, and as it were reflected, fay the altered articulation 
of the inherent vowel of the root. 
The Celtic system of transmuted initials and suppressed suffixes is, how¬ 
ever, subject to one inconvenience, namely, that by tending perhaps towaixls 
a too intimate coalescence of the phonic and logical powers of speech, it may 
be more likely than any method of syntactical expression to obscure in the 
mind of the nation the consciousness of tliuse grannnatical distinctions to 
which it owes its origin. That such lias been the case is evident from ail 
the Welsh and Irish grammars extant; and how then can we wonder at the 
misapplication given to this system by the Teutonic tribes? The manner in 
which I think such a misapplication on their part gave rise to their altered 
scale of articulation is this. 
Those combinations of power, qvfmtity, and form in the mule deviating from 
the radical scale, which in Celtic are hut of syntactical import, and of occa¬ 
sional, although of course most frequent occurrence, were adopted as radical 
and permanent by the Teutonic tribes, who look the tenuis in its altered fonn 
as the basis of a newsede of articulation, radically different from the Sauscritic, 
which they had till tlien retained. This explajiatiou accounts also, as will be 
easily seen, for the second alteration which the new scale underwent in High 
German, the latter taking for a basis the vocal tenuis, whereas the Gothic 
had taken the asjymile, which, as wc have observed, must be considered the 
more ancient of the two forms of alteration. The Gothic, having adopted the 
th (})*) as the short or feeble, and retained tlie d as the long or strong dental 
mule, came to adopt the t as the middle {dental media) i whereas the Old 
High German, having umile that/ its short denial mute and tending to follow 
out this new clmngu by a cnmnlete deviation from the Gothic scale, took the 
th or i as the long, ami the s (ts) as the middle mute.—And perhaps this is 
not the only instance in which ihcTeutotiio mind has been misled, to bestow 
au absolute instead of a relative value on principles derived from the Celtic 
nation. 
The f/nrd. argument in favour of the study of the Celtic, on which we are 
now about to enter, refers to its general Unguistical tearing, as a highly im¬ 
portant member both of the family of human languages in general, and more 
particularly of the so-called Japhetic or Indo-Teutcnic stock. One of the 
grandest results of modern compamtive philology has becu to show, that all 
languages belonging to one slock—and wc may even say, enlarging this 
view, all languages of the earth —are but scattered indications of that primi¬ 
tive state of human intellvet, and more particularly of the imitative faculty, 
under the highest I'xcitcment of poetical iiispiratiou, in which the language 
originated, and with which every language remains conuected as well through 
the physiological unity of the human race as through the historical unity of 
the family to which it more espcuiuUy belongs. Of the divine art by which 
man in that happy primitive state of intidlectual activity was enabled to un- 
deretand the world and himself by means of imitative movements of his 
voice, and, at the same time, of the sacred treasure of ideas thus embodied in 
sonnd with winch he then became entrusted, a certain portion only has been 
preserved and developed by each family of the human race, in accordance 
with its peculiar character aud history, its virtues atid defects. The most 
teautiful portion is undoubtedly that which has fallen to the lot of tlie Ja- 
phetic iamily; but this again has been divided amongst several nations, each 
of which possesses but one dialect of the great Japhetic language, and this but 
* The peculiar somul of the Gothic (and English) > does not form an objection to this 
fact, since this sound i? but Ihc result of a local coalescence of the I with the guttural datiis, 
the latter having accommodated itself to the former by becoming dental. It is by a similar 
process that/lA and kh coalesce and pass respectively into/(p) and ch (x). 
