was son or brother of Ebles Viscount of Thouars, or Arnold, his brother, who was descended from the Counts of Auvergne.

The name of Elie, as a Christian name, in the tenth and succeeding centuries, is frequently met with in the south and southwestern parts of France, and its adoption is almost invariably to be traced to a few families. We have an Elie Comte de Périgord, in 974, which name was borne by nine succeeding Counts of the family of Talleyrand-Périgord, down to the fourteenth century. In the Périgord particularly the name of Elie seems to have been held in high respect. The family of Bourdeilles was one of the most ancient houses of this province, of whom we have an Elie de Bourdeilles, in 1044, and several Elie in succeeding generations. Ebles and Aimery were the names of two grandsons of this Elie. This family was probably a branch of the Elie Viscounts of Limoges, as of course was that of Ségur, of whom there was a Pierre de Ségur in 1295, who had a son Elie. The seigneurs of Perigeux, a most powerful family, and who were styled by Geoffry de Vigeois, who lived in the twelfth century, a race "alti sanguinis," were also known by the surname of Elie, as early as the beginning of the twelfth century. A Pierre de Perigeux lived in 1104. Pierre Helie, and William Helie his father, were benefactors to the abbey of Chancelade, in 1115. Branches took the surname of Plastulf, Arenes, and Prévôt, retaining the Christian names of Pierre and Elie. The following persons were undoubtedly of the same stock :-William Helias, who was vicar of Bourdeaux and lord of L'Isle (which is contiguous to Bourdeilles and Perigeux), in 1138; Giraud Helias, who is witness to a charter of a William Duke of Aquitaine, before 1137; Elias de Malemort, who was Archbishop of Bourdeaux in 1195; and Gerard de Malemort, in 1228.2 These were cousins; the former is styled "antiquæ nobilis que familiæ." Marlemort, or Mortemart, was a castle and seigneurie in the province of La Marche. Alice, daughter and heir of William seigneur de Mortemart, Availles, and St. Germain, married, in 1205, Aimery de Rochechouart.3 The armorial bearings of the preceding families do not contain fleurs de lis; but, as none of their early seals are to be met with, that remark can only apply to their known coats.

1 Courcelles' Dict. de la Noblesse, v. 257, 272, and 346.

2 Gallia Christiana, ii. 274, 282-6-8, 483, App.

3 Desbois, art. Rochechouart.

That

The arms of Pompadour were three towers, and would indicate a connection with the family of the Baron de la Tour d'Auvergne, who was a Crusader in 1096, and who bore gules une tour d'argent, maconnée de sable. Bernard II., seigneur de la Tour d'Auvergne, bore, in 1270, d'azure semée de fleurs de lis d'or, une tour d'argent maconnée de sable. Herbert II., Vicomte de Thouars, a Crusader in 1096, bore or, semée de fleurs de lis d'azure, un franc quartier de gules.

there is an affinity and common origin of them all, has been sufficiently shown. There is one family, however, who bore fleurs de lis, and frequently adopted the Christian name of Elie, which there is little doubt was of the same ancestry as the Elie, Alluye, Ales, &c., viz., that of Fors or Vivonne, a branch of that of Lusignan.1

A coat attributed to Fors is three fleurs de lis. Temp. John occurs an Elias de Fors, in England. De Forz or De Fortibus, Earls of Albemarle, were of this family, though their affiliation is not known. In 1128, an Elie was Seigneur de Vivonne. The arms of Vivonne and Fors, as generally borne, were a chief; this was charged with three fleurs de lis, as borne by a Vivonne, according to the Dict. Gen. (3 vols. 12mo). The Scotch baronetical family of Broun is said to descend from a Norman Baron Le Brun in the twelfth century, descended from the kings of France. Their arms are three fleurs de lis. Hugh the fourth, of Lusignan, was styled Le Brun, and was, doubtless, ancestor of this family. Their first known ancestor is Hugh le Veneur, Seigneur de Lusignan. Hugh the third, his grandson, was living in 967, and in 1010, when he was witness to a charter of his son. Their ancestor is said to be Girard, Count of Auvergne in 839, son-inlaw of Pepin, and progenitor of the Counts of Poictiers. Here we arrive at the same ancestry as those of the Viscounts of Limoges. Gerard and William were favourite names of the Elie, Alises, &c.; they were both of frequent use in the south-western provinces of France; the latter was a favourite name of the Counts of Poictiers and Dukes of Aquitaine. Hugh was the name for several generations, we have seen, of the families of Alluye and Lusignan. As there can be now little doubt that they were both of one lineage, it is important to inquire into the dates and particulars of the divergence of the branches from the main stem. In 862, we find a Stephen, son of Hugh, Count of Auvergne; it is not improbable from one of these might be descended the two families in question, or at least that of Alluye, as we find an Alcher d'Aluia in 1090, who had three sons, Hugh, Walter, and Stephen. Gerard, Vicomte de Limoges in 975, we have had reason to suppose closely related to the first Hugh d'Alluye, who was a contemporary.

It has been said that the family of Lusignan were of the bloodroyal of France (through the Counts of Auvergne). It now remains to show this, and thereby to attempt to establish the proposition with which we set out, viz.: that the Ellises of France and England were descended from an Elias or Louis, of the royal race of the former country.

Girard Count of Auvergne, in 839, was son of Bernard Count 1 Du Chesne's Hist. de la Maison de Chastillon, p. 517; Lusignan and Vivonne are contiguous, and both in Poitou.

of Toulouse, and Viscount of Narbonne, who was son of William, one of the most distinguished generals and nobles of the court. of Charlemagne, of the same royal race, and, moreover, allied to him by marriage. He was loaded by that monarch with honours and dignities: he was Duke of Aquitaine and Gascony; Count of Toulouse, Provence, Languedoc, and Auvergne; and Viscount of Narbonne; and is supposed to be the same person as William-auCornet, Count of Orange, progenitor of the House of Nassau. He retired to a monastery of his own foundation in 806, and died there in 812, being afterwards canonized, and known in history as St. William. He is called by Eginhard "propinquus regis,' and was son of Theodoric, Count of Autun and Duke of Burgundy, who was son of Childebrand, brother of Charles Martel. Theodoric had a brother Nevelon, who was ancestor of the Capetian house of France, and William Count of Blois, who died 834.2 The question now arises-When and by whom was the name of Helias or Elie first used as a surname, and can it be clearly identified with Louis? The name of Louis is not met with earlier than in the person of Louis le Debonnaire, son of Charlemagne ; and the name of Elie or Elias is only twice met with at this early period. An Elias, we have seen, was Bishop of Chartres, 840-9. An Hélie, with others, vassals who possessed the fiefs and tithes of the Abbey of St. Benoite of Dijon, was ordered by Louis le Debonnaire to repair their churches.3 Here we have an Elie in Burgundy, about the middle of the ninth century; he and the bishop could not have been brothers; if first cousins, their ancestor of the same name must have lived a century previous. The extreme paucity of documents at this early period precludes the hope of much further elucidation; but, taking all these circumstances together, there cannot be much doubt that, in these cases, Helias and Elie were synonymous with Louis, Helouis, or Chlovis. Osbert and Osborn, Gerald and Gerard, Arnulph and Arnaud, are synonymous; and it is quite conceivable that similar variations should have been made of the name of Chlovis.

As to the hereditary use of the name as a surname, the Martels, at the time of the Conquest, there is little doubt, inherited that appellative through three centuries, from Charles Martel. In 916, William Count of Angouleme got the cognomen of Taillefer, which was borne through three centuries to Isabella Taillefer, wife of John King of England, and, it is pretty certain, was the original form of Talvase and Taillebois. In the eleventh century,

1 There is an interesting biography of him in Ordericus Vitalis. 2 The authorities for this paragraph, and many other passages, are— Anderson's Royal Genealogies; L'Art de Vérifier les Dates; Moreri's and Anselme's Dictionaries; and Dom Vaissette's Histoire de Languedoc.

3 Du Chesne, Histoire des Ducs de Burgoyne, p. 115; Gallia Christiana, iv. 668.

E

we have innumerable examples of the hereditary use of a Christian name as a surname; and analogy justifies the belief that Elias, or Elie, or Alis, which was used as a surname in the eleventh century, was so used in at least the two previous centuries.

The origin of the device of the fleur de lis has given rise to considerable speculation; but the popular notion of its deriving its name from Louis, and being the exclusive ensign of French dominion, is totally erroneous. Montfaucon has shown that it was not only assumed by the Frankish, but also by the Lombard and other Teutonic princes. In his great work, he gives engravings of statues of the Merovingian and Capetian race of kings, on whose sceptres and crowns the fleur de lis is distinctly represented. But it is to be met with in remains from Babylon, and in sculptures from Nineveh. There is little doubt that the lotus is the flower intended by it; and that, we know, was regarded as of peculiar mystic import in Egypt and throughout the East. In course of time, it seems to have become the settled and peculiar armorial bearings of the kings of France, certainly long before the time of Charlemagne, as the descent of families bearing it from his ancestors will demonstrate.1

Mention has been made of the family of Blois, who bore fleurs de lis. They were descended from Childebrand, son of Pepin. Hugh Comte de Maine, in 1010, is said to be descended of the royal race of France. The family of De Maine, of whom Guy du Maine occurs in 1375, bore semée de fleurs de lis. The Counts of Vexin-Français were descended from Pepin. The family of Pontoise, a cadet, bore azure three fleurs de lis, or, a lambel ermine. The family of Nanteuil were also descended of the same race, and bore gules six fleurs de lis, or three, two, and There are French families named Du Liz and De Lis, who bear fleurs de lis, the particulars of whom, however, are very scanty. The family of St. Liz (Senlis) bore argent, two bars gules, in chief three fleurs de lis argent. A fleur de lis is on the seal of Stephen de Liz, Prior of Lewes in the thirteenth century.3 The fleur de lis is borne by the English family of Lee, perhaps

one.

1 See an elaborate series of articles on the subject in Notes and Queries for the early part of 1856, which comprehend a list of families who bear the fleur de lis in their coat armour: the author, however, commits the mistake of supposing that these were concessions from the kings of France, not admitting the possibility of descent from the early monarchs.

2 Confusion is likely to arise from mistaking three similar but distinct families:-1. Maine, sprung from the Comtes de Maine; 2. Mayenne, from a town so called in Maine; and, 3. Mans, a town in the Vexin-Français, which gave name to Walter or Dreux de Mans, husband of Goda, sister of Edward the Confessor. The three are (in ignorance) indiscriminately called Maine, Mayenne, and Mans.

3 Sussex Archæological Collections, vol. 1.

originally A-lis or De-lis. Sir Arnold Delis was a distinguished friend of Walter de Lacy, temp. William I., and might not im. probably have been an Alis. Three fleurs de lis were the arms of one of the fifteen tribes of Wales, from which sprung the family of Ellis of Ystmyllyn, who bore that coat, but not the name of Ellis, till the period when Welsh surnames became settled. Roderick (? Theodoric) the Great, King of Wales, 843-76, may have been descended from the early kings of France, and originated this coat of arms in Wales; he had a grandson named Elis, who, not improbably, was the prototype of the numerous Lewises and Ellises to be found in that country. The reason why these and certain other Christian names are so common as surnames in Welsh families, is, because, about the time of Henry VIII., the father's Christian name was adopted by the son as a fixed patronymic.2

ADDENDA.

Vol. III. of the Collection des Cartularies de France, edited by M. Guérard, contains the cartulary of the abbey of St. Trinity, at Rouen. This document sheds increased and altogether new light on the origin of the Alises. Page 451, there is a charter of William Fitz-Osborn, dated 1068, which is signed by William his son, and William Alis. If this were the William Alis of Domesday, the speculations in the text, as to his birth and death, are groundless; if his father, his presumed paternity is incorrect. But there is no reason to suppose him to be one or the other, but probably uncle of the Domesday William Alis. The benefaction of William Alis to the priory of St. Denis, is out of his manor of "Aldington" (Auditon): this establishment was founded in 1124, William de Pontearch witnessing the charter; this William Alis must therefore have lived at or after that date; it could hardly be the same person who occurs as a witness, as above, in 1068, but must, doubtless, have been the Domesday William Alis. A William Alis, in the third generation, appears in the person of "William de Alz," a benefactor to Roche Abbey, in Yorkshire, between 1147, the date of its foundation, and 1165, the date of the confirmatory charter of Richard de Busli, reciting the gift. (Dugdale.) Phillip Alis of 1166 had a brother Ralph, mentioned in a charter to the monks of Abergavenny. (Dugdale.)

Grimoldus de Mara occurs (p. 453) as a witness to a charter, apparently about the time of the Conquest. De la Mare is in the canton of Pavilly. Richard Fresnel occurs in Ordericus Vitalis (ii. 191), as mesne tenant, along with William Halis, of the Earl of Brettville. This person belonged

1 He is frequently mentioned in the romance of Fulkes Fitz-Warin, translated by Mr. Wright.

2 At the end of the Dictionnaire Généalogique (3 vols. 12mo), so often quoted, is a list of families who bore fleurs de lis; and there is said to have been a work published in 2 vols. 8vo, in Paris, entirely on the subject of the Fleur de lis.