国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

sénge, lion; ta, horse; tse, top, &c. In general, if two substantives are combined, the
one showing the specific quality precedes; but the specific modification is, as a rule,
the second part, if its form is that of an adjective.

Ancient history or præ-buddhist myths could not be traced in geographical names;
unless some of the superstitious denominations, as Mórdo, the oracle's stones, may be
considered as instances. Not unfrequently a combination of several words is formed
into one name, which then assumes a considerable length, as Sang-gye-chi-ku-
sung-thug-chi-tén (or Himis), the support of the meaning of the Buddha's precepts;
Tsomotethúng, the lake, the wild horse's drinking place.

Hindu names in Tibet occurring in native Indian maps (as also in European maps)
are nearly always of foreign plain Brahmanical origin; the inhabitants themselves and
those of the immediate environs of the respective object have another, a Tibetan name.
The double name of Gaurisánkar and Chingopámari; and, among many others, the more
generally known names of Mansaráur and Tso Máphan, may be alleged as analogous
cases. Celebrated lamaic establishments frequently have a clerical name besides the one
generally in use; as an instance of this I mention the monastery Himis quoted above.

Every alphabetical arrangement of words at once shows, that, as initials, certain
letters and certain combinations of letters are much less frequent in one language
than in others (the system of transliteration chosen naturally concealing the fact to
some extent); and what is still more surprising, letters which appear indispensable in
one group of languages are entirely wanting in their counterpart in another. As a
very remarkable instance, there is the letter "F," which is altogether wanting in the
Sanskrit elements of Hindostáni and in Tibetan. Detailed comparison shows, more-
over, many minor instances of similar nature; short vowels between certain consonants,
particularly mutes and liquids, are not audibly pronounced, and in many of the
native alphabets not written either. In our European languages, also, we have many
similar instances, where frequently the vowels are dropped in speaking, though they
are written.

Tibetan shows still another physiological and ethnographical modification not
unworthy of attention, viz. that the letters used as terminals are very few in number.
Amongst the thirty letters of consonantal character of the Tibetan alphabet ten¹