later one in Sanskrit by Neriosengh. But the
Pehlevi translation, which was made under the auspices of the
Sassanian kings of Persia, served only to show how completely the
literal and grammatical meaning of the Zend-Avesta was lost even at
that time, in the third century after Christ; while the Sanskrit
translation was clearly made, not from the original, but from the
Pehlevi. It is true, also, that even in more modern times the Parsis
of Bombay were able to give to Anquetil Duperron and other Europeans
what they considered as a translation of the Zend-Avesta in modern
Persian. But a scholar like Burnouf, who endeavoured for the first
time to give an account of every word in the Zend text, to explain
each grammatical termination, to parse every sentence, and to
establish the true meaning of each term by an etymological analysis
and by a comparison of cognate words in Sanskrit, was able to derive
but scant assistance from these traditional translations. Professor
Spiegel, to whom we owe a complete edition and translation of the
Zend-Avesta, and who has devoted the whole of his life to the
elucidation of the Zoroastrian religion, attributes a higher value to
the tradition of the Parsis than Dr. Haug. But he also is obliged to
admit that he could ascribe no greater authority to these traditional
translations and glosses than a Biblical scholar might allow to
Rabbinical commentaries. All scholars are agreed in fact on this, that
whether the tradition be right or wrong, it requires in either case to
be confirmed by an independent grammatical and etymological analysis
of the original text. Such an analysis is no doubt as liable to error
as the traditional translation itself, but it possesses this
advantage, that it gives reasons for every word that has to be
translated, and for every sentence that has to be construed. It is an
excellent discipline to the mind even where the results at which we
arrive are doubtful or erroneous, and it has imparted to these studies
a scientific value and general interest which they could not otherwise
have acquired.

We shall give a few specimens of the translations proposed by
different scholars of one or two verses of the Zend-Avesta. We cannot
here enter into the grammatical arguments by which each of these
translations is supported. We only wish to show what is the present
state of Zend scholarship, and though we would by no means disguise
the fact of its somewhat chaotic character, yet