word, to which the Parsis attribute a signification which a comparison of passages and the study of the languages belonging to the same family neither confirm nor explain, to justify the sense given by the Parsis, or to find another. I commenced by detaching from the word to be translated its formative and suffixed terminations, which I had learned from the grammatical analysis of other words, in which the concurrence of Neriosengh, of Anquetil, and of the comparison of languages, left no doubt. I thus reduced the word, about which there was difficulty, to its simplest elements, or what is called its radical. And once master of this radical, I sought to discover whether the languages with which the Zend has most relation, the Sanscrit, Greek, Latin, the Germanic dialects, etc., bore any traces of it. This method led me, in a great number of instances, to very curious results. Thus I have established that the list of Sanscrit roots contained almost all the radicals whose meaning I sought; but that these radicals were infrequently used, if they were used at all in the classical Sanscrit, and that in order to find them in the language it was necessary to ascend to the Vedas. These old radicals were ordinarily strangers to the Greek and Latin languages, for otherwise I would have recognized them more speedily: some were found only in the Germanic dialects. So that the Zend and Sanscrit radicals, viewed with reference to their employment, naturally divided themselves into classes, the most marked of which only I shall indicate at present: 1. Zend radicals, belonging almost exclusively to the language of the Vedas, or to the most ancient Sanscrit, very rare in Greek and Latin, more common in the Germanic languages. 2. Zend radicals not found in classical Sanscrit, but which being mentioned in the lists of roots, have certainly belonged to the language, and probably to its most ancient form: this numerous class is rare in the learned idioms of Europe. 3. Zend radicals belonging to all ages of the Sanscrit, and common to the Greek, Latin, Germanic, Slavic and Celtic tongues. This class is the most numerous of all, and it may be said to form the common fund of all these languages. 4. Finally, Zend radicals, which I have not been able to refer to any known radical of these different languages, but

which I have almost always found again more or less altered in the Persian dictionary.

"If, as I venture to hope, these results at least in the general are incontestable, they cast new light upon the statistics of one of the richest families of human languages. In the first place they establish the high antiquity of the Zend, of which a considerable part is thus found cotemporary with the primitive dialect of the Vedas. In the second place, they evidently prove that the different languages which compose the Sanscritic family should not be regarded as derived one from the other, but that laying aside the different ages of their culture, which establish among them an appearance of chronological succession, they belong primitively to one and the same fund from which they have drawn in unequal proportions. This inequality so striking in the employment of the radicals is found again. in the greater or less development which these radicals have received in the different idioms which have preserved them. Thus a root, which in Sanscrit has remained unproductive, has in Zend given birth to numerous offshoots. Another stopping in the midst of its growth has run through only the first period in one of these idioms, and in another only the last. In a word, whether in derivatives or radicals, nothing is absolutely equal in all these languages, but all set out from an originally common fund and are developed by the same laws.

"This community of origin, of which I met such convincing proofs at every step, emboldened me to attempt an account of a certain number of Zend words, which I saw resist the means of analysis, whose process and results have just been summarily indicated. The comparison of words identical or almost identical in Zend and in Sanscrit, for example, had given me a certain number of laws of permutation of letters; laws, whose certainty is greater, the greater the number of observations upon which they rest, and insofar as they have their ultimate reason in the peculiar constitution of the vocal organs. Zend words which differ from Sanscrit only by the change of one letter, and to which the application of one of these laws could be made with certainty, become the base from which I raised myself to other words, in which the simultaneous application of several laws was necessary. So that I came to explain

Zend words very different in sound from the corresponding Sanscrit terms, and to refer them by the comparative analysis of their elements to the form in which they appear in other idioms. I am far from concealing from myself the inconveniences attached to the exclusive employment of such a method, and I am not ignorant of the dangers of applying it without discretion. For the worth of the rules of permutation is not precisely the same for words which differ completely from each other as for those which are almost alike, and the certainty of these laws decreases in some measure in proportion to the need there is of applying them. But the appreciation of the different circumstances, which can permit or limit their use, belongs to criticism, and I hope it will not be found that I have in this work refused to the reader any of the means of verification which it was my duty to furnish him."

By methods such as this the exhumation of this fossil tongue has been accomplished. And it is a most interesting as well as valuable fact for science, that in philology, as in natural history the fossil remains of what has been extinct for ages, fill chasms and supply missing members in existing species and genera. The recovery of the Zend cannot as yet be considered complete. The general question even is still in dispute among those who have made it their special study, what comparative weight is to be attributed in cases of conflict to traditional aid and to that of the kindred tongues. For some passages of the Avesta, Spiegel does not even venture to propose a translation: in others he speaks with great hesitation and doubt. And when the promised translation of Westergaard appears, who is the champion of dialectic aids as Spiegel of tradition, there will, without doubt, be no small divergence between them. Still the work is essentially done. The language is understood; its structure and general character have been fully exposed; and its relations to the great family of languages within which it is embraced, have been definitely settled.

The Avesta, as we possess it, is a motley jumble of prayers, ritual prescriptions and dogmatic statements, mostly in the form of questions answered by direct address of Ahura-mazda to Zarathustra (Zoroaster.) The Vendidad, the Yaçna, which

is wholly of a liturgical character, and the Vispered, a small collection of invocations, constitute together what is called the Vendidad-Sade. To these are to be added the Yeshts, and a few other ancient fragments. The Bundehesh and other religious writings of the Parsis, manifestly belong to a much later period. Parsi tradition asserts that the books of Zoroaster consisted originally of twenty-one nosks or chapters, and that all which now remains is but a fragment of one of these. The destruction of the remainder is charged upon Alexander the Great, who, after translating all that related to astronomy, medicine, and other sciences into Greek, committed them to the flames; the priests subsequently restoring as much as they could from memory. Spiegel thinks it probable that it was first reduced to writing in its present form, during the Bactrian dominion, in the centuries just before or after the birth of Christ. The second part of the Yaçna, which is the only portion written in measure, is supposed to be the oldest. The Vendidad and the first part of the Yaçna, belong to a somewhat later date and perhaps a different place. The Yeshts are later still. The character in which the manuscripts are now written, is not older than the sixth century of the Christian era. The word Avesta strictly means "text." Zend, though commonly applied to the language since the time of Anquetil, has properly no such meaning, but denotes translation or commentary, and is the designation of the Pehlevi version. Spiegel proposes to abolish it as the name of the language and substitute Old Bactrian.

ART. III.-Religion in America: or, an Account of the Origin, Relation to the State, and present Condition of the Evangelical Churches in the United States. With notices of the Unevangelical Denominations. By Robert Baird. New York: Harper & Brothers. 1856. 8vo. pp. 696.

THIS fine volume is the enlargement of a work which its excellent author published under the same title about twelve years ago, but which he has re-written and greatly improved, by new labour and the results of wider observation. Few persons have had better opportunities than Dr. Baird of knowing exactly what is needed in Europe on the subject of the American Churches. He has, therefore, been enabled to adapt his various studies to the wants of transatlantic readers; and from this point of view his performance must continually be judged. The former edition had a wide circulation in several languages of Europe, and contributed to increase that just esteem in which the author is held by Evangelical Christians in the Old World.

We paid our respects to the volume on its first appearance, and we still adhere to the favourable opinion then expressed. We still hold, that there is no American, however well informed, who may not read it with instruction, and refer to it as a cyclopedia of facts not elsewhere extant in connection. This is the more true, when we reflect that the members of different ecclesiastical bodies live very much apart, so that a Lutheran and a Methodist often know less of one another, than either knows of his own fellowship in England or Prussia. We still rejoice that the work has fallen, not into the hands of a partisan or a fanatic, but of a mild, generous, and large-minded man, who has done as much as any other towards the promotion of fraternity among differing sects. Since we thus expressed ourselves, the work has undergone a thorough revision, for the sake of which the author has retired from important public employments. It has been brought down in its details to our own times. Every sentence has been read, and almost every enumeration has been changed; as was inevitable in regard to a country and a period like our own.

Of eight books, the First is devoted to preliminary remarks,