purpose; and admits in a very sparing degree of any modification of the primitive. The formation, also, of four tenses of the verb by means of particles is peculiar to itself; and it partakes of the Arabic grammatical system, while it differs from the Sanscrit, in affixing parts only of the pronouns to the noun or verb with which they are placed in construction. The Persian, at the same time, is dissimilar from the Sanscrit, in having neither dual number nor genders, and in its adjectives being indeclinable.

But no conceivable cause can be assigned for such radical differences, had the grammatical structure of Persian ever been the same as that of Sanscrit; because experience sufficiently proves that conquest alone can effect any material change in language, and that even its influence is not powerful enough to produce a complete alteration in the grammatical forms to which a people has been long accustomed. As, therefore, there is no indication in tradition or history that a nation speaking Sanscrit ever conquered Persia, it must be admitted that its grammatical structure is alone sufficient to demonstrate that Persian is not indebted to that language for its origin. But the Sanscrit words which are still discoverable in Persian are much too numerous, and expressive of too great a diversity of ideas, to sanction the supposition, that they could have been introduced into it by mere intercourse, whether hostile or commercial, between the Persians and a people speaking Sanscrit.

So far, therefore, as relates to the Hindus, these remarks of Sir W. Jones would appear to be well founded, "So that the three families, whose lineage we have examined in former discourses, had left visible traces of themselves in Iran [Persia] long before the Tartars and Arabs had rushed from their deserts, and returned to that very country, from which, in all probability, they originally proceeded, and which the Hindus had abandoned in an earlier age, with positive commands from their legislators to revisit it no more. I close this head with observing, that no supposition of a mere political or commercial intercourse between the different nations will account for the Sanscrit and Chaldaic words, which we find in

the old Persian tongues: because they are, in the first place, too numerous to have been introduced by such means; and, secondly, are not the names of exotic animals, commodities, or arts, but those of material elements, parts of the body, natural objects and relations, affections of the mind, and other ideas common to the whole race of man.' ""* Had, however, the original inhabitants of Persia been Hindus, the people who remained in it must have spoken precisely the same language as those who migrated from it, and the colony must either have retained this language, or adopted a new one. In the first case, consequently, Persian ought even at this day to contain a greater number of Sanscrit words, and to exhibit a grammatical system nearly similar to that of Sanscrit; and, in the other case, though Sanscrit might retain many terms common to Persian, it ought at the same time to exhibit distinctly its mixed origin: but, on the contrary, Sanscrit is the purest of languages, as it does not contain a single exotic word, and, while the Sanscrit grammatical system is easily identified in Greek, not a trace of it can be discovered in Persian.

That part, therefore, of Sir W. Jones's hypothesis which supposes that the aborigines of Persia were Hindus is untenable; but it is equally evident that a people speaking Sanscrit must have at some time not only inhabited this country, but have also possessed such influence in it as could have occasioned the introduction of so many words of their own tongue into the vernacular language. As, also, fifty-five of the Sanscrit words found in Persian are equally found in Greek, it must necessarily follow that they had passed into Persian one or two centuries before the poems of Homer were written, because at that time the Greek language appears to have been completely formed. But there is no indication in history, or in Sanscrit works, that the Hindus ever made any foreign conquests; and the contrary would appear most probable, from the great antiquity of those institutions which prevent a Hindu from leaving for any cause the land of holiness. The establishment in Persia, therefore, of a people

* Sir W. Jones's Works, vol. i. p. 83.

speaking Sanscrit must evidently have preceded their entrance into India, and, if not aborigines of the country, they must necessarily have immigrated into it from some other kingdom. Thus, again, the conjecture irresistibly presents itself, that this people speaking Sanscrit could be no other than a numerous colony which had migrated from Babylon on its conquest by the Ninus of Herodotus, part of which established itself in Persia and part proceeded on to

India.

Nor, if this conjecture be admitted, can it seem improbable, from the wide-spread fame of the Chaldeans, that this colony should be enabled to improve the Persians in arts and civilisation, and thus to occasion the introduction of many Babylonian or Sanscrit words into the language of Persia: for the similar introduction of Latin words into all the dialects of Celtic now existing, and of Sanscrit into the vernacular dialects of India, sufficiently shows that the conquest of a country is not the only means by which its language may become affected by foreign influence. It may however, be objected that, in these instances, this influence prevailed in consequence of a new religion having been propagated in the foreign language; and that the universal voice of antiquity attests that the religion of Persia was totally dissimilar from that of India. But it must be recollected that the earliest writer who has given a description of the Persians flourished so late as 450 years B. C., and, consequently, that his authority cannot determine what the popular faith of the Persians may have been 800 years before his time. All accounts, also, ancient, Parsi, and Muhammadan, concur in ascribing to Zoroaster, or Zardusht, the introduction of a new religion into Persia. Hence, it may be reasonably concluded, from the systems of belief that existed in the neighbouring countries, that the popular faith subverted by Zoroaster was idolatry; and that his great merit must have consisted in withdrawing the Persians from the worship of idols, and in imparting to them juster notions of the Supreme Being.

Although, also, the religion of Babylonia was no doubt idolatry, this colony might have introduced, as in India, various alterations into the

system of popular faith which might have then prevailed in Persia. The memory even of one remarkable circumstance, the institution of caste, has been preserved by Muhammadan writers, which identifies the ancient Persian religion with that of the Egyptians and Hindus, and thus renders its common origin almost demonstrated. For Tabari, in his account of Jemshid, relates, that "he divided the people into four classes, one consisted of soldiers, another of learned men, another of scribes and artizans, and another of agriculturists: and he commanded each class to follow their respective occupations, the agriculturists to reside in the country, the scribes to exercise the office of magistrates, the soldiers to attend at his gate, and he placed the learned men over the three other classes, and commanded them to take care that each class pursued its own occupation." These words evidently show that this description depends not on any account of a similar institution in India which Tabari might have heard of*, but must have been derived from some tradition preserved in Persia. No other traces, however, of the ancient religion of the Persians previous to Zoroaster can now be discovered. But this tradition, supported by the irrefutable testimony of language, must tend to render it highly probable that a colony, similar in all respects to that which introduced the Brahminical religion into India, was also about the same time established in Persia, and that both these colonies proceeded from one and the same country, the ancient Babylonia.

But, in whatever manner the cause of the existence of Sanscrit words in Persian may be explained, it is undeniable that except them no other foreign terms can be found in this language; and, consequently, its purity and originality demonstrate that neither Scythians, Celts, Pelasgi, or Goths ever inhabited Persia. It hence, also, appears that the words in Persian which seem to be cognate with terms in Greek, Latin, and the Teutonic dialects, have been principally derived

* It is to be remarked that Tabari was born in A. D. 838, and that the Muhammadans never made any successful attack on India until Mahmud of Ghoznin in A. D. 1000. Nor does there appear to have existed any intercourse between India and Bagdad, which could have enabled Tabari to acquire any knowledge of the institutions of the Hindus.

from one common origin, the Sanscrit; and that the few* which cannot be traced to this source are not sufficiently numerous to invalidate this conclusion; because neither geography, chronology, nor history warrants the supposition that they could ever have passed from these languages into Persian, though it is not possible to point out the manner in which they may have passed from the latter into the former. The existence, at the same time, of 265 Sanscrit words in Persian, most fully evinces that Pahlvi could not have been the common speech of Persia at the time of the Arabian conquest; for, after that event, the state of the country rendered the introduction into Persian of so many Sanscrit words expressive of such a diversity of ideas utterly impossible. Nor, when the antiquity of the Hindu institutions is considered, does the coexistence of three distinct languages in Persia, and the introduction of such numerous foreign terms into one of these only, appear in the least more probable. To suppose, indeed, a colony, so powerful as to occasion so many words of its own tongue to have passed into the vernacular language of the whole of Persia, to have been established in the province of Fars and its dependencies only, at least 1200 years B. C., is an opinion much too absurd to be maintained by any person. These Sanscrit words, therefore, and the remote period at which they must have been introduced into Persian, must alone be sufficient to demonstrate that the people of Persia have always spoken but one and the same mother tongue; and, consequently, affinity of language, the most indisputable of testimonies, completely disproves the supposition that Persian is the same language that was spoken by the Scythians, from which it has been conjectured that the Celtic, Pelasgic, and Gothic have been derived.

* I have inserted in the preceding Chapter such as I have been able to discover, amounting to forty-one in number.