First Appearance in European History.— From whatever cause, it is certain that a confusion did exist between the ’Aτσiγκαvοι and Αθίγγανοι, which renders it extremely difficult to determine whether the Byzantine historians are speaking of Gipsies or heretics in seven passages collected by Miklosich. It appears from these that Αθίγγανοι, described as magicians, soothsayers, and serpent-charmers, first emerge in Byzantine history under Nicephorus I. (802-11), were banished by Michael I. (811-13), and were restored to favour by Michael II. (820-29); but Miklosich’s reasons for absolutely identifying them with Gipsies, and positively asserting the latter to have appeared at Byzantium in 810 under Nicephorus, are hard to recognize. Less dubious seems an extract from the Georgian Life of Giorgi Mtharsmindel (1 lth century), which describes how at Constantinople certain descendants of the race of Simon Magus, Atsinkan by name, sorcerers and famous rogues, slew wild beasts by their magic arts in the presence of Bagrat IV. Such passages are open to some doubt; hardly so the following from the Itinerarium Symoηis Simeonis (ed. by J. Nasmith, Camb. 1778), where-Fitz Simeon, a Franciscan friar of Dublin, describing his stay in Crete in 1322, says “We there saw a people living outside the city (of Candia), who worship according to the Greek rite, and declare themselves of the race of Ham. They rarely or never stop in one place more than thirty days, but, as though accursed of heaven, wander from field to field with little, oblong, black, low tents, like those of the Arabs, or from cave to cave.” The empress Catherine de Valois, again, who died in 1346, granted to the suzerains of Corfu authority to reduce to vassalage certain homines vageniti coming from the mainland, who under the Venetians formed in 1386 the nucleus of a feudum Acinganorum that lasted down to the present century. About 1378 the Venetian governor of Nauplion confirmed the Acingani of that Greek colony in privileges granted by his predecessors; and in 1387 Mircea I., waiwode of Wallachia, renewed a grant made by his uncle Ladislaus to the monastery of St Anthony at Voditza of forty salaschi (tents) of Acigani. Other documents might be cited, but these are enough to show that in the 14th century Gipsies existed in the Balkan peninsula and islands of the Levant; that in Wallachia they were reduced to a state of bondage (from which they were only freed in 1856); and that nowhere were they regarded as new-comers, so that by these documents it is impossible to fix the date of the first Gipsy immigration. More than this, a metrical German paraphrase of Genesis, made by an Austrian monk about 1122, preserved at Vienna, and edited by Hoffmann in his Fundgruben für Geschichte deutscher Sprache (Breslau, 1837), goes far to prove that Gipsies were known in Austria three centuries before the commonly-accepted date of their appearance in that country. A passage relating to Hagar’s descendants (Gen. xvi. 15) runs :—“So she (Hagar) had this son; they named him Ishmael. It is from him the Ishmaelites descend. They journey far through the world; we call them chaltsmide (lit. cold-smiths). . . . They have no house nor country; everywhere they are found alike; they wander over the country, abusing people by their knaveries. Thus they deceive men,—robbing no one openly.” That here by chaltsmide, Ishmaelites, and descendants of Hagar Gipsies are meant, scarcely admits of doubt, seeing that the smith’s is still the Gipsies’ leading handicraft; that Lusignan in 1573 speaks of the Gipsies of Cyprus as “Cinquanes, otherwise called Ag ariens f and that in German and Danish Rotwalsch or thieves’ slang Geschmeilim and Smaelem (i.e., Ishmaelites) signify “Gipsies.” The κωμοδρόμοι also of Byzantine writers were possibly Gipsies, being defined by Ducange as “circulatores atque adeo Fabri aerarii qui pér pagos cursitant: ut hodie passim apud nos, quos Chaudronniers dicimus.” Theophanes (758-818) speaks under the date 544 of a κωμ.οδρόμ.ος from Italy.

Later Movements. — Late in 1417 there came to Lüneburg a band of 300 wanderers, “black as Tartars and calling themselves SecaniT At their head rode a “duke” and “count,” splendidly dressed, and leading like nobles dogs of chase; next came a motley crew afoot; and women and children brought up the rear in waggons. They bore among other letters of safe-conduct one granted by the emperor Sigismund, and professed themselves engaged on a seven years’ pilgrimage, imposed by their bishops in expiation of apostacy from the Christian faith. From Lüneburg they passed to Hamburg, Lübeck, Wismar, Rostock, Stralsund, and Greifswald, camping by night outside the walls, thieving by day, “wherefore several were taken and slain” (cf. the contemporary annals of Korner and Rufus, and Krantz’s Saxonia, 1520). In 1418 they journeyed southwards through Meissen, Leipsic, and Hesse, and, entering Switzerland, arrived at Zurich on August 31st, visiting also Basel, Bern, and Solothurn, according to Conrad Justinger (died 1426), who speaks of them as “more than 200 baptized heathens from Egypt.” They now split up into two bands, the first of which appeared before Augsburg (November 1, 1418), the second before Sisteron in Provence (October 1, 1419), where the terrified citizens bestowed on the “Saracens” a hundred loaves. Next comes a long notice of a troop of fully 100 lean, black, hideous Egyptians in the Chronica di Bologna (July 18, 1422), which tells how the sorceress, “Duke” Andrew’s wife, could read the past and future of men’s lives; but Bologna in fifteen days became too hot forthem, so by way of Forli—where “certain said they were from India”—the pilgrims travelled on to Rome. Their object was to procure fresh letters from the pope; and such they afterwards produced, though of their sojourn in the imperial city no record has yet been published. To the burghers of Ratisbon Gipsies presented themselves in 1424; they pitched their tents again before its walls in 1426; and at Paris in 1427 the fair of Landit was attended by a duke, a count, and ten other mounted pilgrims, late renegades of “Lower Egypt,” whose women practised palmistry and cleared everybody’s pockets. Later we hear of Gipsies at Arnheim (1429), at Metz (1430), at Erfurt (1432), and in Bavaria (1433),—these and all notices of the seventeen years preceding referring probably to the movements of a single ubiquitous band, sent forward to spy out the lands of promise, and composed of from 600 to 1400 persons. For not until 1438 did the great tide of westward immigration begin to flow; then, not in hundreds but thousands, headed no longer by paltry “dukes” and “counts,” but by a “king,” King Zindl, the Gipsies poured over Germany, Italy, and France, reaching Poland by 1501, Sweden by 1512, and having already appeared in Spain in 1447. We find them in England in 1514 (A Dyaloge of Syr Thomas More, 1529), but nothing is known of the date of their landing; and in Scotland the earliest certain record of their presence is an entry in the books of the Lord High Treasurer: “Apr. 22, 1505. Item, to the Egyptianis, be the kingis command, vij lib.” (Pitcairn’s Criminal Trials, Edin. 1833, vol. iii. p. 592). In a “King of Rowmais” ( '? Romas, Gipsies), twice mentioned in entries of July 1492, as also in the “Erie of Grece” (1502), “King Cristall” (1530), and the “King of Cipre” (1532), one dimly recognizes four Gipsy chiefs: and with Gipsies perhaps the Saracens may be identified, whom a tradition represents as making depredations in Scotland prior to 1460 (Simson, p. 98). In no other country were Gipsies better received than in this, where they “dansit before the king in Halyrudhous” (1530); where James IV. gave (July 5, 1505) Anthonius Gagino, count of Little Egypt, a letter of commendation to the king of Denmark; and where James V. subscribed a writ (February 15, 1540) in favour of “oure louit Johnne Faw, lord and erle of Litill Egipt,” to whose son and successor he granted authority to hang and punish all “Egyptians” within the realm (May 26, 1540). But in 1541 an Act was passed, commanding the “Egyptians to pass forth of the realm” under pain of death, and similar edicts were issued before and afterwards in most of the European states —Germany (1497), Spain (1499), France (1504), England (1531), Denmark (1536), Moravia (1538), Poland (1557),

&c. Conveying across the seas was among the milder measures adopted; it is, however, noteworthy as one of the causes of the dispersion of the tribes. Under Henry VIII. Gipsies were shipped from England to Norway (Wright’s History of Ludlow, pp. 389-92) or France; whilst by the latter power, so lately as 1802, the bands infesting Bayonne and Mauléon were caught by night as in a net, huddled on shipboard, and landed on the coast of Africa (Michel, Pays Basque, p. 137). In Scotland four Faas were hanged at Edinburgh in 1611 “for abyding within the kingdome, they being Egiptianis;” and in 1636 doom was pronounced on other “Egyptians” at Haddington, the “men to be hangit, and the weomen to be drowned; and suche of the weomen as hes children to be scourgit throw the burgh and brunt in the cheeke.” Under the English statute of 1562 (repealed 1783) making· it felony without benefit of clergy to be merely seen for a month in the fellowship of Gipsies, five men were hanged at Durham “for being Egyptians,” 8th August 1592. Still greater were the cruelties and injustice suffered by Gipsies on the Continent, since there, to the charge of kidnapping, were added the weightier imputations of being cannibals and emissaries of the Turk. Quinones recounts how in 1629 four Estremaduran Gitanos owned under torture to having eaten a friar, a pilgrim, and a woman of their tribe; and in 1782 forty-five Hungarian Gipsies were· beheaded, quartered, or hanged on a like monstrous charge. First racked till they confessed the crime of murder, they were brought to the spot where their victims were said to be buried, and when no bodies appeared they were racked again. “We ate them” was their despairing cry; and forthwith the journals teemed with accounts of “eighty-five persons roasted by Gipsy cannibals”; straightway the “cannibals” were hurried to the scaffold. Then Joseph ∙∏. sent a commission down, whose inquiries showed that no one had been murdered—except the victims of the false accusation. The full, impartial annals of the race have still to be compiled, from edicts and law-books, from local histories and a few monographs like Dirk’s- Geschiedkuηdige onderzoekingen aangaande het verblijf der Heidens of Egyptiers in de Noordelijke Nederland en (Utrecht, 1850), or Weber’s Zigeuner in Sachsen, 1488— 1792, in vol. ii. of his Aus vier Jahrhunderten (Leip. 1861).

Language.— Until lately the information about the Gipsy language to be gathered from books was meagre in the extreme. The thirteen works published prior to 1840 which furnish specimens of the Anglo-Romani dialect— Boorde (1547), Bryant (1785), Bright (1818), Copsey (1818), Harriot (1830), Roberts (1836), <fcc.—together contain but 396 genuine stems, besides 69 doubtful words, and furnish scarcely any examples of the grammar. Nor are the Continental works cited by Pott, from Vulcanius (1589) downwards, much more copious. Even to-day there are still great gaps in our knowledge, especially of the dialects outside of Europe; but enough has been done to show that from the Nile to the Arctic Ocean, from the Euphrates to the Atlantic, the Gipsies speak, with dialectal variations, one and the selfsame speech. The Romani names for “water,” “fire,” “hair,” and “eye,’* are in Persia pdni, aik, bdl, and aki ; in Egypt pàni, dg, bal, and ankhi∙, in Norway pani, jag, bal, and jak ; in England pani, yog, bal, and yok. And these four instances, which might be multiplied indefinitely, serve further to show, by their resemblance to the Hindi pamï, ag, bal, and ankh, that in Romani we have an Indian tongue. Rüdiger first compared Romani, so long regarded as a thieves’ jargon, withone of the New Indian dialects, and in 1782 published the result of the comparison in his Neuester Zuwachs der Sprachkunde. In 1783 Grellmann’s Historischer Versuch reaped all the fruits of Rüdiger’s research; and in the same year Marsden was independently led to a like discovery. The conclusion that the Gipsies wandered forth from India is now almost universally accepted, but when, or from what part of India, are questions on which few have done more than idly speculate. Whether Romani is derived from Hindi, Marathi, &c., can only be determined by minute investigations, which, long neglected, arc now being carried on by various Orientalists. They have at least established that Romani stands in the relation of a sister, not a daughter, to the seven principal New Indian dialects. Its forms are often more primitive than theirs, sometimes than those of Pali or the Prakrits, e.g., vast, “hand” (Sanskrit hasta, Pali hattha}, kasht, “wood” (Sanskrit köshlha, Pali kattha),vusht, “lip” (Sanskrit oshtha, Pali ottha), trash, “fear” (Sanskrit trasa, Pali tas), trin, “three” (Sanskrit tri,trvηi∙, Pali ti, tinι), and pral, “brother” (Sanskrit bhrata, Pali bhata). And while the archaisms of Romani forbid us to derive it from Hindi or Marathi, some of its seemingly modern forms are the result of independent development. On the other hand, our knowledge of Romani itself, and of the multitudinous spoken dialects of India, is not at present sufficient to warrant our pronouncing the former more primitive than any of them; and as a fact many of its archaisms may be paralleled in the languages of Dardistan and Kafiristan (cf. Miklosich, Beiträge, iv. 45-54). Thus there are difficulties on both sides in the way of adjudicating between the opinions of Ascoli and Miklosich. The former maintains (Saggi Iηdiani, vol. ii., 1875) that Romani, preserving certain consonantal nexus which had almost entirely disappeared at the epoch of the most ancient Prakrit texts, approaches Sanskrit more nearly even than Pali—conclusions, he observes, that harmonize well with Bataillard’s pre-historic theory. Miklosich, on the other hand, opposes that theory in Meyer’s Konversations-Lexikon (3d ed. 1878), where he infers from the agreement of Romani in its phonetic laws and system of case-endings with the modern Aryan languages of India that the emigration cannot have taken place till these were formed, i.e., until after the Prakrit period.

In Europe Miklosich distinguishes thirteen Romani dialects—the Greek or Turkish, Roumanian, Hungarian, Moravo-Bohemian, German, Polo-Lithuanian, Russian, Finnish, Scandinavian, Anglo-Scottish, Italian, Basque, and Spanish. To these should be added the Welsh, which, generally unintelligible to the English Gipsy, is one of the most perfect, as it has also been the least studied, of all the dialects. As a general rule, the further these dialects remove from Turkey, the more corrupt have they become, so that the Gipsies of Spain, of Scandinavia, and in great measure of England, know no case or verb endings other than those of the lands of their adoption. From Turkish Romani, therefore, and Welsh the following examples will be drawn. The Turkish (marked T.) are taken from Paspati; the Welsh (W.) are derived from letters and stories written by John Roberts, the oldest living harper, whose thorough knowledge of his language is probably unique.

The definite article, wanting in Asia, is supplied in every European dialect by the Greek <5 and ή — ό for the masculine, ή for the feminine and the oblique cases, c.g., W. Botchdαs ow bearcηgaro e vaver tringcngey, “the sailor asked the other three,” E colley pendas, “the Gipsy woman said.” The indefinite article, in some dialects supplied by yek, “one,” is still omitted by the Welsh and the “deeper” English Gipsies, e.g., Yckerporro gougco ta porrey gougeyjiυenes undra borrow veshestay, “once (an) old man and (an) old woman were living in (a) great wood.” Romani has no trace of either a dual number or a neuter gender. Excepting monosyllables, most of its nouns terminate in ∙o (masc.) and -i (fem.), as raklo, “lad,” rakli, “girl.” Masculine nouns ending in a consonant form their feminines in -ni, as rom, “husband,” romni, “wife.” Inanimate objects are indifferently masculine or feminine: to the former belonging gav, “town,” and gad, “shirt;” to the latter nok, “nose,” and bok, a hunger.” Rom, “a husband,” and rakli, i, a girl,” are thus declined in Turkish Romani [table]

Here the so-called genitive is in reality an adjective. It precedes and agrees in gender with its noun, c.g., T. c derviséskeri rakli, “the dervish’s daughter;” W. sonekaicsko mochto, “a golden box,” and dakey pen, “mother’s sister." Welsh Gipsies often use the dative...