, analysed the language of the Avesta scientifically. He
proved--

1. That Zend was not a corrupted Sanskrit, as supposed by W.
Erskine, but that it differed from it as Greek, Latin, or
Lithuanian differed from one another and from Sanskrit.

2. That the modern Persian was really derived from Zend as
Italian was from Latin; and

3. That the Avesta, or the works of Zoroaster, must have
been reduced to writing at least previously to Alexander's
conquest. The opinion that Zend was an artificial language
(an opinion held by men of great eminence in Oriental
philology, beginning with Sir W. Jones) is passed over by
Rask as not deserving of refutation.

The first edition of the Zend texts, the critical restitution of the
MSS., the outlines of a Zend grammar, with the translation and
philological anatomy of considerable portions of the Zoroastrian
writings, were the work of the late Eugene Burnouf. He was the real
founder of Zend philology. It is clear from his works, and from Bopp's
valuable remarks in his 'Comparative Grammar,' that Zend in its
grammar and dictionary is nearer to Sanskrit than any other
Indo-European language. Many Zend words can be retranslated into
Sanskrit simply by changing the Zend letters into their corresponding
forms in Sanskrit. With regard to the Correspondence of Letters in
Grimm's sense of the word, Zend ranges with Sanskrit and the classical
languages. It differs from Sanskrit principally in its sibilants,
nasals, and aspirates. The Sanskrit s, for instance, is represented by
the Zend h, a change analogous to that of an original s into the
Greek aspirate, only that in Greek this change is not general. Thus
the geographical name hapta hendu, which occurs in the Avesta, becomes
intelligible if we retranslate the Zend h into the Sanskrit s. For
sapta sindhu, or the Seven Rivers, is the old Vaidik name of India
itself, derived from the five rivers of the Penjab, together with the
Indus, and the Sarasvati.

Where Sanskrit differs in words or grammatical peculiarities from the
northern members of the Aryan family, it frequently coincides with
Zend. The numerals are the same in all these languages up to 100. The
name for thousand, however, sahasra, is peculiar to Sanskrit, and does
not occur in any of the Indo-European dialects except in Zend, where
it becomes haza_n_ra. In the same manner the German and Slavonic
languages have a word for th