(September 6, 1641), but took refuge in Flanders, with the Duc de Guise, against whom a similar sentence had been pronounced, and remained there until the death of Richelieu, followed by that of Louis XIII., left him at liberty to return to France. When, in 1647, Guise went to Naples, to endeavour to exploit the revolt of Masaniello to his own advantage, Modène accompanied him and greatly distinguished himself. He was eventually, however, taken prisoner by the Spaniards and held captive until 1650. On his return to France, he meddled no more with public affairs, but occupied himself with the care of his neglected estates and in the compilation of a valuable history of the revolution in Naples, reprinted, in 1826, under the title of Mémoires du Comte de Modène. It is to be noted here that from the early autumn of 1641 until the summer of 1643 the Comte de Modène was absent from France.

Some time in the early weeks of the year 1643, probably either in the last week in February or the first in March, Madeleine's father, Joseph Béjart the elder, died; and on March 10, Marie Hervé, his widow, presented herself before the Civil Lieutenant of Paris, where, in the name, and as guardian, of Joseph, Madeleine, Geneviève, Louis, and "a little girl not yet baptized," children under age (i.e. under twenty-five) of the said deceased and herself, she represented that "the inheritance of her deceased husband being charged with heavy debts without any property wherewith to acquit them, she feared that it would be more burdensome than profitable," and, accordingly, declared her intention of renouncing it. Her request was supported by her brother-in-law, Pierre Béjart, procureur to the Châtelet, and other relatives, and on June 10 of the same year she was permitted to make the renunciation she desired.

Now who was this "little unbaptized girl"? Without a shadow of doubt, Armande Béjart, the future wife of Molière; on this point all the poet's biographers are unanimous. Was she, as represented, the daughter of Marie Hervé? That is the question which has afforded material for a controversy which has already lasted for nearly two hundred and fifty years and seems not unlikely to continue till the end of all things, for the most fantastic theories, for a small library of books and pamphlets, and for review and newspaper articles without number. For some see in this little girl a sister, others a daughter of Madeleine Béjart, and the truth is of the most vital importance to the honour of the great man whose wife Armande became.

That the latter impression was almost universal amongst Molière's contemporaries is beyond question, nor is the fact one that need occasion any surprise. Every one, that is to say, every one connected with, or interested in, the theatrical world, was aware that, early in life, Madeleine Béjart had had a little girl; while, on the other hand, the birth of Marie Hervé's child, which was of no public interest, and which, moreover, probably took place not in Paris, but in one of the adjacent villages,[4] was known to very few. A young girl grew up with Madeleine, who was tenderly attached to her; it was Armande; but gossip confounded her with the child Francoise, of whom all trace seems to have been lost, and the wiseacres smiled the smile begotten of superior knowledge when any stranger to Paris chanced to refer to the girl as Madeleine's sister.

For over a century and a half this belief remained unchallenged. Hostile or sympathetic, all who wrote of Molière—La Grange, Grimarest, Breuze de la Martinière, Bayle, Donneau de Visé—shared the common opinion in regard to the origin of Armande Béjart. In 1821, however, there was quite a flutter of excitement in literary circles, for in that year Beffara discovered Molière's acte de mariage, in which Armande is spoken of as the daughter of Joseph Béjart and his widow, Marie Hervé. Forty-two years later, the old scandal, which in the interim had been partly revived by M. Fournier (Études sur la vie et les œuvres de Molière) and M. Bazin (Notes historiques sur Molière), received another severe blow by Eudore Soulié's discovery of the deed of March 10, 1643, already mentioned, wherein Marie Hervé requested permission to renounce the succession to her husband's property, and which confirmed the statement made in the acte de mariage. Such evidence, one would naturally suppose, would have been accepted as conclusive, and the matter set at rest once and for all. But tradition dies hard; not a few Molièristes refused to renounce an opinion sanctioned by so many generations, and M. Jules Loiseleur, a writer who enjoyed a considerable, and not undeserved, reputation as an unraveller of historical mysteries, propounded, on behalf of his fellow-sceptics, the following theory.

The declarations made by Marie Hervé, in the deed of March 10, 1643, and again in the acte de mariage, that Armande was her child, were, he maintains, deliberate falsehoods, conceived in the interests of her daughter, Madeleine. At the beginning of the year 1643, Madeleine was about to become a mother, for the second time, not, of course, by the Comte de Modène, who had been in exile for nearly two years, but by some new lover. Fearing that if Modène returned and learned the fact, he would refuse to resume the liaison, which she hoped might one day be regularised (M. Loiseleur was under the impression that Madame de Modène was dead, whereas she lived until 1649), she begged her mother to recognise the child as her own; a request to which that complacent old lady, whose husband was just dead, or on the point of death, readily consented.

Now this ingenious theory is based on the advanced age of Marie Hervé—she was then about fifty-three—and the belief that she had not had a child since the birth of Louis Béjart, afterwards a prominent member of Molière's troupe, who was born on November 14 or 15, 1630, that is to say, more than twelve years earlier, which facts rendered it highly improbable that she could have been the mother of Armande; and M. Loiseleur supports his contention by pointing out that the two eldest children, Joseph and Madeleine, described in the deed of March 10, 1643, as minors, were over twenty-five, and that their age was purposely understated to make their mother appear younger than she was, and so facilitate the fraud. This point has been contested by Mr. Andrew Lang, in his admirable article on Molière in the Encyclopædia Britannica, but is really of no importance, as if M. Loiseleur had exercised a little more care, he would have found that so far from more than twelve years having elapsed between the birth of the last of Marie Hervé's children and that of Armande, she had had a little girl less than three and a half years before (November 30, 1639), baptized, in the parish of Saint-Sauveur, by the name of Bénigne Madeleine, the second name being doubtless intended as a compliment to Madeleine Béjart, who acted as marraine.[5] Whereby M. Loiseleur's argument disappears, and his theory with it.

Nevertheless, the fact remains that Armande's contemporaries saw in her not a sister, but a daughter of Madeleine Béjart, and, with this belief, they held another, to wit, that Molière had been, previous to his marriage with the younger sister, the lover of the elder. From which two suppositions sprang one of the most hideous accusations that has ever sullied the reputation of a great man.

Molière, like most successful men, had a good many enemies, and was