In the beginning of the Vendidad, or first of the Parsi collection of sacred books, known as the Zendavesta, we are told that the supreme deity Ahura-Mazda created a country full of delights, but difficult of access, and the name of this country was Aryana Vaëjo. So charming was this primitive country that, had it not been made difficult of approach, the whole animate creation would have flocked thither and quite overwhelmed it. But this state of things did not long continue; for Ahriman, or Anramainyus, the spirit of darkness, was the implacable adversary of Ormuzd, or Ahura-Mazda, the spirit of light, and took pleasure in spoiling all his creations. So this death-dealing enemy, with the aid of his daëvas, or demons, created a great serpent and brought ten months of winter cold upon the land, so that Aryana Vaëjo was no longer a comfortable dwelling-place. The good spirit then created a new home for his people, called Sugdha; but the adversary spoiled this by creating a kind of wasp which devastated the fields and brought death to the cattle. Then Ahura-Mazda made a third habitat, which was called the high and holy Mum; but the dark demon now whispered evil reports and stirred up strife, until here, too, life became unendurable, and the beautiful land of Bakhdhi, or Baktria, was created as a fourth home for the children of light. So the warfare went on, until no less than sixteen countries are enumerated as successively created and made uncomfortable. In the last region of all the complaint is again of cold weather and hoar-frost; but perhaps in comparison with all the other plagues this now seemed endurable. At all events, the account here ends, with the admission that there are also other regions and places besides those described; as much as to say that we are not here concerned, however, with the history of all mankind.

The book from which this legend is cited is one of the oldest in the literature of the world. It belongs to a more primitive age than the Homeric poems, and may probably be regarded as contemporary with the oldest hymns of the Veda. Written not in the court language of ancient Persia, but in the closely-related archaic dialect of Baktria, — very much as the ecclesiastical services of Russia to-day are written in Old Bulgarian, — the Zendavesta was, in the time of Darius Hystaspes, the sacred book of the most prominent nation in the world. For eleven hundred years afterward the worship of Ahura-Mazda retained its ascendency in the countries between the Euphrates and the Indus, until in the seventh century after Christ this whole region was overrun by Mohammedans, and converted to their faith. For a long time, no doubt, the Magian religion continued to survive alongside of Islam, as we see from the frequent allusions to “fire-worshipers” in the Arabian Nights, where they are indeed most abominably slandered. But after a while the good Ahura-Mazda, yielding to this last and gravest mischief wrought by the adversary, devised yet another abode for the remnant of his people, and led them to Bombay and its neighborhood, where, under the name of “Parsis,” or “Persians,” they still keep up their old ceremonies and traditions.

The legend of the sixteen countries created by the good spirit was regarded by Bunsen as a historical tradition of the migrations by which the ancestors of the Indo-Persians reached the countries.

where, at the beginning of authentic history, we find their descendants. But it will not do to attach too much historical value to legends like this. For, however venerable may be the record, the very mist of antiquity which shrouds it prevents us from knowing how or whence it got the information which it imparts. The story before us, indeed, has neither the pretensions nor the credentials of an authentic historical narrative. It relates long-past events as ascertained not through the sifting of previous human testimony, but by direct revelation from the good spirit to his prophet Zarathustra or Zoroaster. Nevertheless, the geographical succession of the various places mentioned in this legend is very suggestive. With the exception of Aryana Vaëjo, every one of the sixteen abodes seems to be described by a genuine geographical name, though two or three have not yet been satisfactorily determined. Thus Sugdha, the second country, is what the ancients knew as Sogdiana; Mum appears to be the modern Merv, or Margiana; and Baktria, the next in order, has been already mentioned. And so, curiously enough, by stringing together the whole series of names, there is indicated a continuous migration from the region beyond the Oxus, at first southwesterly, and then southeasterly, down to what we now call the Punjab, or “country of five rivers,” but which in the Vedic hymns is somewhat more comprehensively termed the Sapta-Sindhavas, or “Seven Rivers,” and which in our Zend legend is described in identical language as the Hapta Hendu. This larger designation is reached by including, along with the five rivers of the Punjab, the Sarasvati and the Indus, or “The River,” par excellence. Having thus reached the northwestern confines of Hindustan, in the fifteenth country created by Ahura-Mazda, the legend here informs us that Anramainyus devised “untimely evils and unbearable heat;” and thereupon we are abruptly transported, in the sixteenth region, to the cool neighborhood of the Caspian Sea, perhaps the country of the Medes.

Now, however difficult it may be to accept such an account as properly historical, the course of migration here indicated is so thoroughly in accordance with all that we know of the relations between the peoples of the Persian Empire and the dominant race of Hindus in India that it is hard not to grant to it some traditionary value. It would appear, at least, that when the Vendidad was composed the worshipers of Ahura-Mazda must have believed that their ancestors came from somewhere beyond the Oxus, and traveled in the direction of Hindustan, until something occurred which turned them westward again. This would seem to be the only sound meaning that can be extracted from the legend. But this is in wonderful accordance with the results of modern critical inquiry. From a minute survey of the languages and legends of this whole region, it has been well established that the dominant race in ancient Persia and in ancient India was one and the same; that it approached India from the northwest; and that a great religious schism was accompanied by the westward migration of a large part of the community, while the other part proceeded onward, and established itself in Hindustan. A comparison of the Zendavesta with the Veda—so strongly alike as they are, both in thought and in expression—shows clearly that the occasion of this schism must have been the promulgation of the worship of Ahura-Mazda.

In illustration of this community of origin between the Vedic and Zendavestan peoples, let us refer to the name of the first country which the supreme deity created, — the name of Aryana Vaëjo. This, as already hinted, is not a geographical name. There is no identifiable locality which has ever been called Aryana Vaëjo. The name means simply the starting-place of the Aryans. In later Persian mythology, as represented in the Minokhired, the name came to stand for a terrestrial paradise, where men live for three hundred years, without pain or sickness, where no lies are told, and where ten men eat of one loaf and grow fat thereon. In the Vendidad, however, Aryana Vaëjo is simply the primeval dwelling-place, whatever it may have been, from which the Aryans passed into Sogdiana. Now “Aryan” was the name by which the ancient Persians and the ancient Hindus alike described themselves. In the Vedic hymns the dominant people of India habitually speak of themselves as Aryans, in contrast with the Dasyus, or inferior races of Hindustan, whom they had subdued. Just in the same way Darius Hystaspes, in the inscription upon his tomb, declares himself to be an Aryan, of Aryan descent. The Medes are always called Aryans by Armenian writers; and Herodotos was also familiar with this appellation. In a more special sense the countries between India and Persia, now known as Afghanistan and Cabul, were known throughout classic antiquity as Ariana. Along with this community of name there was close community of speech among these peoples. The court language of the Medes and Persians, as preserved in the cuneiform inscriptions of Darius, the Zend or Baktrian language, in which the sacred books of Zarathustra are written, and the Sanskrit of the Vedic hymns are as clearly dialects of the same parental language as French, Spanish, and Italian are dialects of Latin. These outline facts are all that we need for the present to show how Aryan was the common name for a race which, advancing from the north, acquired supremacy over all the country between the Euphrates and the mouth of the Ganges. Whence these people originally came it would be idle to inquire, but we may fairly conclude that they first attained to something like world-historic importance in the high-lands of Central Asia, somewhere about the sources of the Oxus and the Jaxartes; and this region we regard as “Aryana Vaëjo,” or the most aboriginal spot to which we are able to trace the Aryan people.

We have next to inquire into the meaning of the word Aryan; and this is not a difficult matter, or one about which there is much question. In Sanskrit the word arya, with a short initial a, is applied to cultivators of the soil, and it would seem to be connected etymologically with the Latin arare and the archaic English ear, “to plow.” As men who had risen to an agricultural stage of civilization, the Aryans might no doubt fairly contrast themselves with their nomadic Turanian neighbors, who—as Huns, Tatars, and Turks—have at different times disturbed the Indo-European world, But for the real source of the word, as applied to the race, we must look further. This word arya, “a cultivator of the soil,” came naturally enough in Sanskrit to mean a householder or land-owner, and hence it is not strange that we find it reoccurring, with a long initial a, as an adjective, meaning “noble” or “of good family.” As a national appellative, whether in Sanskrit or Zend, this initial a is always long, and there can be no doubt that the Aryans gave themselves this title as being the noble, aristocratic, or ruling race, in contradistinction to the aboriginal races which they brought into servitude. In this sense of noble, the word frequently occurs in the composition of Persian proper names, such as Ariobarzanes, Ariaramnes, and Ariarathes; just as in old English we have the equivalent word ethel, or noble, in such names as Ethelwolf and Ethelred. As an ethnic name, therefore, the word Aryan seems to have a tinge of patriotic or clannish self-satisfaction about it. But we shall find, I think, that such a shade of meaning has been more than justified by history; for we have now reached a point where we may profitably enlarge the scope of our discussion, and show how the term Aryan is properly applicable, not merely over an Indo-Persian, but over an Indo-European area, comprehending the most dominant races known to history, — the Greeks and Romans, Slays and Teutons, with the highly-composite English, whose language and civilization are now spreading themselves with unexampled rapidity over all the hitherto unoccupied regions of the earth, which the Vendidad did not care or did not know how to specify. In order to explain in what sense we may all properly be called Aryans, we must consider for a moment some of the striking results which have been obtained, within the present century, from the comparative study of languages.

No event of modern times has exerted a more profound and manifold influence upon the intellectual culture of mankind than the English conquest of India. The enlargement of our mental horizon which has resulted therefrom is not less remarkable than that which attended the revival of Greek studies in the fifteenth century. It is not simply that observation of India is making us acquainted with an enormous multitude of primitive social, linguistic, and religious phenomena which formerly were hidden from our notice. In contemplating these phenomena, we have become possessed of a method of study which has already wrought such wonders as to vie with the ointment of the Arabian dervise, that enabled its owner to detect all the buried treasures of the earth. This mighty talisman is the Comparative Method, or the attempt to interpret a fact by comparing it with a series of similar facts, which different circumstances have caused to vary in different degrees. I do not mean to imply that mankind have not always used this method more or less, both in matters of science and in matters of every-day life. Nor do I mean to claim for modern philology any exclusive title to the honor of having shown what can be done by studying phenomena in this way. I do not forget that the classification of living and extinct animals by Cuvier, with reference to palæontological epochs, was a gigantic act of comparison, which first made it possible for us to understand the past history of life on our globe. It is none the less true not only that the systematic employment of the comparative method on an extensive scale is the most notable philosophic achievement of the nineteenth century, but also that its first great triumph was the establishment of the Aryan, or Indo-European, family of languages. This triumph was prepared by the study of Sanskrit, which ensued upon the English conquest of India. Previous to this, indeed, the close resemblance between Greek and Latin had been often enough remarked, and theories had been entertained concerning a primeval kinship between the peoples of Greece and Italy. But in the case of peoples so similar in aspect and so closely connected with one another from time immemorial, this similarity of speech did not provoke much curiosity. It was quite otherwise when a language unmistakably akin to Greek and Latin, both in grammar and vocabulary, was discovered in such an out-of-the-way country as Hindustan, and among a people who had hitherto been generally supposed to be barbarians. The discovery was emphasized by the fact that no such obvious resemblances existed in Hebrew, a language much nearer geographically and historically, and from which there had been no end of futile attempts to derive Latin and Greek. Further interest was excited when it became known that this newly-found language contained an enormous mass of literature alleged to be the oldest in the world. All things thus combined to stimulate speculation as to the true character of the relationship between Sanskrit and the languages of Greece and Rome. This relationship was not one of parentage. It has been a common popular error to suppose that Latin and Greek are derived from Sanskrit; but from the first no such view was countenanced by competent scholars. About 1790, Sir William Jones declared his opinion that the three languages were sprung from “some common source, which perhaps no longer exists.” Persian also he was inclined to attribute to the same source, and he hinted at the possibility that Gothic and Keltic might be included in the group. This was coming very near to the conception of an Indo-European family of languages. But that conception was not clearly formed until nearly twenty years later, and then it was reached not by a great philological scholar, but by a poet and literary critic. In 1808, Friedrich Schlegel maintained that the languages of India, Persia, Greece, Italy, and Germany were connected by common descent from an extinct language, just as the modern Romanic languages are connected by common descent from Latin; and for the whole family he proposed the name Indo-Germanic. The correctness of this view was demonstrated by Bopp, in his Comparative Grammar, published from 1833 to 1852, in which the Zend, Armenian, Slavonic, and Lithuanian languages also were added to the group. The Keltic languages were included about the same time, and the name Indo-Germanic was extended to Indo-European. Within the last fifteen years—mainly through the influence of Max Müller’s writings—the name Aryan has come into general use as the most convenient designation of the whole family. The use of the word in this extensive sense has indeed been objected to by Professor Whitney and others, who urge that it is properly applicable only to the Indo-Persian branch of the family; and in strictness their argument seems to be sound enough. There is no evidence that any of the European peoples have ever called themselves Aryans, and the traces of the name which Müller has sought to point out in Europe are very scanty and obscure. According to Stephanus of Byzantium, Aria was an old name for Thrace, and among the ancient Germans we find a tribe of Arii and such proper names as Ariovistus; but it is by no means certain that these names are in any way connected with the original Arya. Nor did Pictet meet with any better success in his attempt to find Arya in the name of Erin or Ireland, the home of the Eri, or Irish. This modern name is a contracted form. Its root in old Keltic seems to have been Iver, which is the same as the Sanskrit avara, “western.” It appears in the Latin Avernus, a famous lake on the west coast of Italy, as well as in Ivernia, or Hibernia, the western island. This old word Iver has been shortened to Ir or Er, and out of this, by putting on their own terminations, the English have made Ire-land, the home of the Ir-ish, or “westerners.” But in spite of the fact that we find no certain traces of the name Aryan in the European languages, I believe that the modern use of the word, as descriptive of the whole family, is likely to prevail. It is a much less cumbrous term than “Indo-European,” and, while it is advantageously free from geographical restrictions, it emphasizes, at the same time, the fundamental fact that the Aryana Vaëjo, or prehistoric starting-point of the eastern members of the family, was also the starting-point of the western members. It implies—what every one admits to be true—that the dominant race in Europe came from Central Asia. And, still further, it serves admirably as a name for the extinct mother tongue from which all the Indo-European languages have descended. By many scholars this primitive tongue is itself called Indo-European; but I am unable to see any propriety in giving such a name to a language which, as being confessedly spoken north of the Oxus and east of the Caspian, was certainly neither Indian nor European in any sense. It seems to me much better, and more in conformity to the general style of philologists, to call this ancestral language “Old Aryan,” just as we say “Old Norse” for the primitive form of Danish, Swedish, and Norwegian.

As we now proceed to take a brief survey of the Aryan domain, I think we shall realize the advantage of having a word that is independent of geographical limits. The Aryana of the present day is much more than an Indo-European region. Its eastern boundaries have altered but little for many centuries; but on the west it has extended to the Pacific coast of America, and on the other side of the world it has begun to annex territory in South Africa and Australia. Indeed, if we are to judge from what has been going on since the times of Drake and Frobisher, it seems in every way likely that men of English speech will by and by have seized upon every part of the earth’s surface not already covered by a well-established civilization, and will have converted them all into Aryan countries. But our linguistic term Aryan is independent of such changes. Since prehistoric times eight principal divisions of Aryan speech have existed, but these groups of languages have had very different careers, and some of them are rapidly becoming extinct. The first great separation of Aryan tribes was the separation between the invaders of Indo-Persia and the invaders of Europe. We have already observed how the language of the Indo-Persians became divided in twain. In the Indic class of languages, comprising the classical Sanskrit, the Prakrit of later dramatic writers, the Pali, or sacred language of the Buddhists in Ceylon, and some twenty modern dialects spoken chiefly in the northern half of Hindustan, we have the first grand division of Aryan speech. The second or Iranic class comprehends the Zend, the ancient Persian of the cuneiform inscriptions, the Parsi of Bombay, the Pushtu of Afghanistan, modern Persian, Armenian, Kurdish, and the Ossetian spoken in the Caucasus. Concerning these two grand divisions, we need only observe that the extremely close resemblance between Sanskrit and Zend would seem to indicate that the separation of the two occurred at a comparatively late date, though it would perhaps be difficult to suppose it later than two thousand years before Christ. Long before this time western tribes of Aryans must have crossed the Volga and begun the conquest of Europe. First appear to have come the Kelts, whose languages constitute the third great division. These languages diverge considerably from the common type, and were the latest to be recognized as Aryan in character, — a fact which is quite in harmony with the opinion that they were the first to branch off from the original stock. The Kelts have always been an important race, but their languages have not thriven in the world. Keltic geographical names are scattered all over Europe, and in the eastern part such words as Dnieper, Don, and Danube testify to the former presence of the language, in which don was a common name for water or river. The Kelts formed a large part of the populations of Spain and Northern Italy, and a principal part of the populations of Gaul and Britain, when these countries were subjected to Roman dominion; and as late as the Christian era they were to be found in large numbers as far east as Bohemia. Since then they have been partly conquered and partly driven westward by Romans and Teutons, without ceasing to be conspicuous as a race; but their languages have sunk into comparative obscurity, and are fast disappearing. The Gauls, who showed such a remarkable aptitude for taking on the manners of their conquerors that by the fourth century their country was almost as thoroughly Romanized as Italy itself, forgot their own language with wonderful ease. It was so completely trampled out by Latin that very scanty vestiges remain to show what it was, if we except geographical names. At the present day two groups of Keltic languages remain: the Gaelic, still spoken in Scotland, Ireland, and the Isle of Man; and the Kymric, or old British, which survives in Welsh and in the dialect of Brittany. A third dialect of Kymric was formerly spoken in Cornwall, but it died in 1770 with Dame Dolly Dentreath.

Concerning the fourth and fifth grand divisions of Aryan speech—the Italic and Hellenic—but little need be said. These languages are too illustrious to stand in need of much description. The relationship between them is closer than in the case of any other Aryan languages of different class, save the Zend and Sanskrit; and this close resemblance justifies the inference that the separation between Greeks and Italians was comparatively recent. They would appear to have entered Europe somewhat later than the Kelts, but everything connected with their prehistoric career is extremely problematical. To the Hellenic class belong only two languages, — the uncultivated Albanian and the Greek, which was stereotyped so early and so thoroughly by literary culture that to the Athenian school-boy of to-day the history of Herodotos can hardly seem written in a foreign tongue. To the Italic class belong the ancient Umbrian and Oscan and the Latin, which still survives under the variously modified forms of Italian, French, Provençal, Spanish, Portuguese, Rumansch, and Wallachian. To the linguist the history of these Romanic dialects is peculiarly valuable, as illustrating, with the aid of plentiful documents, a process of divergence somewhat similar to that which previously broke up the Old Aryan into different languages.

The Teutons, whose languages form our sixth grand division, seem to have entered Europe after the tribes already mentioned. About Caesar’s time we find Teutons driving Kelts out of Germany, and threatening invasions into Gaul; but during most of classic antiquity the centre of Teutonism seems to have been farther east than Germany. The greater part of what is now European Turkey was occupied by Goths in the time of Herodotos, and for eight centuries afterwards. The ancient Thracians were Goths, according to Grimm, and so were the Getæ. And since the Christian era Teutonic tribes appeared in what is now Southern Russia. The terrible irruption of non-Aryan Huns from Asia, in the fifth century, drove these tribes westward, and brought them into collision with the Empire. Of the Gothic language nothing remains save a portion of a translation of the Bible, made by Ulfilas in the fourth century. The other branches of Teutonic speech—Scandinavian, High German, and Low German, of which our own English is the most important dialect—are too well known to require comment.

The seventh and eighth grand divisions of Aryan language are the closely-related Lettic and Slavonic. The Lettic languages, like the Keltic, are fast dying out. “Old Prussian, which has been dead for two centuries, is only represented by the Catechism of Albert of Brandenburg.” Lettish and Lithuanian, of which the latter is remarkable for its strong resemblance to Sanskrit, are still spoken in the Baltic provinces of Russia.

As for the Slavs, they appear in history north of the Black Sea about the time of Trajan, and begin to be frequently mentioned in the sixth century. Since then they have pushed westward far into the Teutonic domain, but have nowhere, save in Russia, retained political independence. Of the fifteen or more Slavonic languages, the Old Bulgarian and the modern Russian, Polish, Bohemian, Croatian, and Serbian are of most importance.

Looking thus over our modern linguistic Aryana, we see that in the Old World it pretty nearly covers the geographical area included between the Ganges and the Atlantic Ocean. Small regions of non-Aryan speech, however, occur here and there within this area, and a brief glance at these will serve to increase the definiteness of our knowledge.

Wherever non-Aryan languages are spoken within this Indo-European domain, it is for either one of two reasons. Such languages are spoken either by descendants of the aboriginal tribes, whom the invading Aryans overcame, or by descendants of non-Aryan invaders, who have pushed in at a later date, and secured for themselves a lodgment upon Aryan soil. Of the first class we find a few sporadic instances. The language variously called the Bask, Euskarian, or Iberian, now spoken in the Asturias and about the Pyrenees, has no similarity whatever to the Aryan languages. It is spoken by the scanty remnant of a people who in immemorial antiquity seem to have been spread all over Western Europe, but who were for the most part conquered or absorbed by the Keltic van of the Aryan invasion. The case may have been similar with the Iapygian and Etruscan, which were long ago trampled out in Italy by the Latin; but on this obscure point I would hardly venture an opinion. In Northern Europe, Finnish, Esthonian, and Lappish are still spoken by races pushed into the corner by Teutons and Slavs. A perfect Babel of aboriginal dialects still exists in the inaccessible fastnesses of the Caucasus; and many of the high. lands of India similarly shelter primitive non-Aryan tribes, whose forefathers refused to submit to Brahmanic oppression. It is a characteristic of such remnants of conquered speech to subsist only in out-of-the-way or undesirable corners. On the other band, Turkish and Hungarian are foreign tongues brought into the Indo-European area by recent intruders. Both these languages belong to the Altaic, Turanian, or Tataric family, spoken by nomadic tribes all over Northern Asia, and including in Europe the Finnish and its congeners above mentioned. The Hungarian has especially strong affinities with the Finnish, while the nearest relatives to Turkish are to be found about Khiva and Bokhara, in the Tataric region which Russia is so rapidly subjugating.

We have now arrived at a tolerably correct idea of what is meant by the word Aryan. But one important point must not be overlooked. In its modern sense we have seen that the word is a linguistic term. It describes community of language. As we now use the word, Aryans are people who speak Aryan, or Indo-European, languages. It is only in a secondary way that this word can be used as an ethnological term, describing community of race. We are so accustomed to consider language a mark of race that it is difficult to avoid using linguistic epithets in an ethnological sense, and a good deal of confused thinking sometimes results from this. We have above alluded to the Aryans as a dominant race, which long since overran Europe and is now spreading over America; yet it is easy to see that we have no means of determining how far the various peoples who speak Aryan languages are of common descent. It is never safe to use language as a direct criterion of race, for speech and blood depend on different sets of circumstances, which do not always vary together. We of the English race have much Keltic blood in our veins, but very few Keltisms in our speech; while, on the other hand, with a vocabulary nearly half made up of Latin words, we have either no Roman blood in our veins, or so little as not to be worth mentioning. During the past twenty-five years Frenchmen have had a good deal to say about the “Latin race.” There could hardly be a more flagrant instance of the perversion of a linguistic name to ethnological purposes. In reality, even in Caesar’s time, the dominant tribes of Latium had become well-nigh absorbed in the non-Latin, though kindred, Italic races which had succumbed to them. After Gaul had been conquered, it learned Roman manners, but without receiving any very large infusion of Roman blood. In point of race the French are Kelts, with a considerable sub-stratum of Iberian and super-stratum of Teutonic blood, — the former chiefly in the south, the latter chiefly in the north. Between Frenchmen, Spaniards, and Northern Italians there is, indeed, a close ethnic affinity; but this is because they are all to a great extent Kelts, not because they have all learned to speak dialects of Latin.

Now if we pursue the matter a little farther, and inquire what we mean by saying that these three peoples are in great part Keltic, we shall find that a similar qualification is needed. Obviously, we mean that they are Keltic in so far as they are descended from people who formerly spoke Keltic languages. Our knowledge of the prehistoric career of the Kelts is too small to admit of our meaning more than this. In just the same way, when we say that Spaniards and Englishmen and Russians are akin to each other as being Aryans, we can only mean that they, are in great part descended from people who spoke Aryan languages.

There can be little doubt, however, that all races which have long wandered and fought have become composite to a degree past deciphering. And, however mixed may have been the blood of the Aryan-speaking invaders of Europe, it remains undeniable that the possession of a common language by such great multitudes of people implies a very long period of time, during which their careers must have been moulded by circumstances in common. It implies common habits of thought and a common civilization, such as it was. And this inference is fully confirmed by a comparative study of the myths and superstitions, as well as of the primitive legal ideas and social customs, of the various parts of the Indo-European world. For this reason I think we are justified in speaking of the Aryan race just as we speak, without error, of the English race, though we know that many race elements have combined their energies in the great work of English civilization. I do not say, either, that we may not fairly speak of a Latin race, provided we bear in mind the limitations of the phrase; the objection is not so much to the phrase as to the loose way in which it is customarily used and the absurd inferences which are often grounded on it.

The ethnologist, who deals with skulls and statures and complexions, may venture much farther, sometimes, than the linguist, — though perhaps the greater length of his excursions may not always compensate for their comparative insecurity. It is quite open to the ethnologist to hold that the successive Aryan swarms which colonized Europe were like each other in physiological characteristics, as well as in language and general culture. Differences of complexion, when well marked, are among the most conspicuous differences which distinguish individuals, groups, or races from one another; and they are, moreover, apt to be correlated with deep-seated physiological differences of temperament. In all countries peopled by Europeans there are to be found two contrasted complexions, the blonde and brunette; endlessly complicated and varied by intermarriage, but nevertheless in their extreme examples so strikingly different that a stranger might well be excused for considering them as marks of difference in race. In populations that have long been stationary and isolated from foreign intrusion we do not find such differences of complexion. We do not find them in China or Japan, or among the Samoyeds, or Kafirs, or Pacific islanders, or among the Arabs. It appears to be only among the Indo-European nations that they occur side by side in the same community, as an every-day matter. Now we may account for this coexistence and inter-mingling of contrasted complexions by supposing that the various peoples of Europe have arisen from the intermixing in various proportions of a race that was entirely blonde with a race that was entirely brunette. We know that the Bask or Iberian race, which once seems to have possessed a great part of Europe, was, and still is, uniformly dark complexioned. We may, accordingly, suppose that the Aryan-speaking invaders were uniformly light. The effect of the earlier invasions of Kelts, Italians, and Greeks would be to crowd the dark-skinned Iberians into the three southern peninsulas, into Western Gaul, and into the British Isles. The next step would be the conquest of all these regions, followed by extensive inter-marriage and the general adoption of Aryan speech. In the remotest corner of all, cooped up between the Pyrenees and the Bay of Biscay, — here, if anywhere, a remnant of the aboriginal population might preserve its purity of race and its primitive speech. As a result of these proceedings, the Aryan-speaking peoples of Greece, Italy, Spain, Gaul, and Britain would show a mixture of light and dark complexions, and wherever the invaders had been much less numerous than the aborigines the brunettes would predominate. But now, where the later swarms of Teutons and Slavs came pouring in, the case would have been somewhat altered for them. Their conquerings and interminglings would take place not with a pure-blooded race of dark aborigines, but with the mixed race which had resulted from the foregoing events. One consequence would be an increased percentage of fair complexions in western countries overrun by Teutons, especially in England, Northern France, and Northern Italy. Another consequence would be the partial darkening of Teutons and Slays by intermixture with Kelto-Iberian predecessors in Southern Germany and Austria. Wherever, on the other hand, the new-comers were left pretty much to themselves, as in Northern Germany, Central Russia, and Scandinavia, we should find the auburn hair and blue eyes of the old Aryan still in the ascendent.

For my own part, I am quite inclined to accept this very ingenious hypothesis, which is defended by such a cautious ethnologist as Professor Huxley, and which makes such historic and philological data as we have account remarkably well for the actual distribution of light and dark complexions¹ throughout Europe. It agrees so well with the facts before us that we can hardly do better than adopt it as a provisional explanation, subject to such revision and amendment as may turn out to be necessary. But if we thus admit the existence of a primitive Aryan race that was physically homogeneous, it must be remembered that we admit it on very different grounds from those on which were based the demonstration of a primitive homogeneous Aryan language. The original community of language is a point on which we have reached absolute certainty; the community of race, in any other sense than that of long-continued community of language and culture, is merely a matter of speculation.

Concerning the people and the series of historic events of which Aryana Vaëjo was the legendary starting-point, we have thus obtained much interesting and trustworthy information by the aid of the comparative method of inquiry. For be it observed that the results so far set down have been reached, for the most part, by a mere comparative survey of the various regions of the linguistic and ethnical field with which we have been called upon to deal. We have in this way obtained quite an accurate conception of what is meant when we speak of the Aryans. But as yet we have dealt only with the veriest rudiments of the subject. Nor have we as yet gone far toward illustrating the vast and rich resources of the comparative method. To be able to depict the prehistoric culture of the Aryan-speaking people, to interpret their mythical conceptions, and to unfold the other remarkable truths that lie latent in the variety of their speech, — this is indeed a fruitful achievement. But to show how this has been brought about requires a separate and more detailed form of exposition.