ON THE COMPARISON OF ASIATIC LANGUAGES. 207 
that nearly half my life has been spent in foreign lands, and 
among primitive peoples, and that I have been forced by 
circumstances to acquire the speech of those with whom I 
dwelt—for eight years among Italian peasants, for six among 
Arabs and Turks, and for one year among Kaffres and 
Hottentots, in regions only since that time incorporated in 
our Empire. ‘The study of antiquity, at the same time, has 
obliged me to enquire into the dead languages of Asia; and 
practical knowledge of the vulgar dialects has shown me, 
as It has shown others, that languages are older than their 
written grammars, and. that the archaic speech of peasants 
is more nervous, more simple, and more symbolic, than are 
the polished phrases of literary authors, and of the later 
standard style. But at the same time the absolute importance 
of recognising the distinctions, in grammar and in sound, 
which now divide the great groups from each other, is only 
the more forcibly impressed on the mind by hearing ie 
actual conversation of various races. 
And first as regards sounds. The distinction of sounds 
nearly akin increases with increase of civilisation, and with 
increased delicacy of ear. The scientific alphabets of to-day 
distinguish no less than 86 sounds, including 27 vowels; but 
the oldest Semitic alphabets, rich as they are in sounds 
hardly distinguished by an European ear, are limited to 
22 letters ; and the oldest inscriptions in these oe no note 
of the short vowel sounds. The Akkadian was only 
accustomed to mark 17 sounds in writing, and when the 
Greeks used the Cypriote syllabary they had to content 
themselves with 14 sounds. With this we may contrast the 
alphabets of their descendants, the Turks, having 32 letters 
against the 17 in Akkadian, the Arabs 28 against 22 in 
Hebrew, and the writers of Sanskrit no less than 50 against 
the 14 of the Cypriote syllabary. Nor was this small amount 
of distinction due to want of graphic power, for the symbols 
have decreased steadily in number, while the precision of 
distinction has increased, and the additional letters are very 
generally distinguished from the older only by an added dot 
or line. * The distinctions are also, in very many cases, only 
marked in literature, and not clearly discernable in the 
speech of the ignorant, so that one of two kindred sounds 
becomes characteristic of one dialect, and another takes its 
place in a second dialect of the same language. It is on this 
peculiarity that the comparative study of European languages 
rests, aS on a secure basis; and it has become more and 
