to a close, and the arrangement of any ransom was a slow process, involving much patient bargaining. Hassan refused to accept less than five hundred gold ducats for his slave; the available funds fell short of this amount, and the balance was collected from the Christian traders of Algiers. Cervantes was already embarked for Constantinople when the money was paid on the 19th of September 1580. The first use that he made of his liberty was to cause affidavits of his proceedings at Algiers to be drawn up; he sailed for Spain towards the end of October, landed at Denia in November, and made his way to Madrid. He signed an information before a notary in that city on the 18th of December 1580.
These dates prove that he cannot, as is often alleged, have served under Alva in the Portuguese campaign of 1580: that campaign ended with the battle of Alcántara on the 25th of August 1580. It seems certain, however, that he visited Portugal soon after his return from Algiers, and in May 1581 he was sent from Thomar on a mission to Oran. Construed literally, a formal statement of his services, signed by Cervantes on the 21st of May 1590, makes it appear that he served in the Azores campaigns of 1582–83; but the wording of the document is involved, the claims of Cervantes are confused with those of his brother Rodrigo (who was promoted ensign at the Azores), and on the whole it is doubtful if he took part in either of the expeditions under Santa Cruz. In any case, the stories of his residence in Portugal, and of his love affairs with a noble Portuguese lady who bore him a daughter, are simple inventions. From 1582–3 to 1587 Cervantes seems to have written copiously for the stage, and in the Adjunta al Parnaso he mentions several of his plays as “worthy of praise”; these were Los Tratos de Argel, La Numancia, La Gran Turquesa, La Batalla naval, La Jerusalem, La Amaranta ó la de Mayo, El Bosque amoroso, La Unica y Bizarra Ársinda—“and many others which I do not remember, but that which I most prize and pique myself on was, and is, one called La Confusa which, with all respect to as many sword-and-cloak plays as have been staged up to the present, may take a prominent place as being good among the best.” Of these only Los Tratos de Argel (or El Trato de Argel) and La Numancia have survived, and, though La Numancia contains many fine rhetorical passages, both plays go to prove that the author’s genius was not essentially dramatic. In February 1584 he obtained a licence to print a pastoral novel entitled Primera parte de la Galatea, the copyright of which he sold on the 14th of June to Blas de Robles, a bookseller at Alcalá de Henares, for 1336 reales. On the 12th of December he married Catalina de Palacios Salazar y Vozmediano of Esquivias, eighteen years his junior. The Galatea was published in the spring of 1585, and is frequently said to relate the story of Cervantes’ courtship, and to introduce various distinguished writers under pastoral names. These assertions must be received with great reserve. The birth of an illegitimate daughter, borne to Cervantes by a certain Ana Francisca de Rojas, is referred to 1584, and earlier in that same year the Galatea had passed the censor; with few exceptions, the identifications of the characters in the book with personages in real life are purely conjectural. These circumstances, together with the internal evidence of the work, point to the conclusion that the Galatea was begun and completed before 1583. It was only twice reprinted—once at Lisbon (1590), and once at Paris (1611)—during the author’s lifetime; but it won him a measure of repute, it was his favourite among his books, and during the thirty years that remained to him he repeatedly announced the second part which is promised conditionally in the text. However, it is not greatly to be regretted that the continuation was never published; though the Galatea is interesting as the first deliberate bid for fame on the part of a great genius, it is an exercise in the pseudo-classic literature introduced into Italy by Sannazaro, and transplanted to Spain by the Portuguese Montemõr; and, ingenious or eloquent as the Renaissance prose-pastoral may be, its innate artificiality stifles Cervantes’ rich and glowing realism. He himself recognized its defects; with all his weakness for the Galatea, he ruefully allows that “it proposes something and concludes nothing.” Its comparative failure was a serious matter for Cervantes who had no other resource but his pen; his plays were probably less successful than his account of them would imply, and at any rate play-writing was not at this time a lucrative occupation in Spain. No doubt the death of his father on the 13th of June 1585 increased the burden of Cervantes’ responsibilities; and the dowry of his wife, as appears from a document dated the 9th of August 1586, consisted of nothing more valuable than five vines, an orchard, some household furniture, four beehives, forty-five hens and chickens, one cock and a crucible.
It had become evident that Cervantes could not gain his bread by literature, and in 1587 he went to Seville to seek employment in connexion with the provisioning of the Invincible Armada. He was placed under the orders of Antonio de Guevara, and before the 24th of February was excommunicated for excessive zeal in collecting wheat at Écija. During the next few months he was engaged in gathering stores at Seville and the adjacent district, and after the defeat of the Armada he was retained as commissary to the galleys. Tired of the drudgery, and without any prospect of advancement, on the 21st of May 1590 Cervantes drew up a petition to the king, recording his services and applying for one of four posts then vacant in the American colonies: a place in the department of public accounts in New Granada, the governorship of Soconusco in Guatemala, the position of auditor to the galleys at Cartagena, or that of corregidor in the city of La Paz. The petition was referred to the Council of the Indies, and was annotated with the words:—“Let him look for something nearer home.” Cervantes perforce remained at his post; the work was hard, uncongenial and ill-paid, and the salary was in constant arrears. In November 1590 he was in such straits that he borrowed money to buy himself a suit of clothes, and in August 1592 his sureties were called upon to make good a deficiency of 795 reales in his accounts. His thoughts turned to literature once more, and on the 5th of September 1592, he signed a contract with Rodrigo Osorio undertaking to write six plays at fifty ducats each, no payment to be made unless Osorio considered that each of these pieces was “one of the best ever produced in Spain.” Nothing came of this agreement, and it appears that, between the date of signing it and the 19th of September, Cervantes was imprisoned (for reasons unknown to us) at Castro del Río. He was speedily released, and continued to perquisition as before in Andalusia; but his literary ambitions were not dead, and in May 1595 he won the first prize—three silver spoons—at a poetical tourney held in honour of St Hyacinth at Saragossa. Shortly afterwards Cervantes found himself in difficulties with the exchequer officials. He entrusted a sum of 7400 reales to a merchant named Simón Freire de Lima with instructions to pay the amount into the treasury at Madrid; the agent became bankrupt and absconded, leaving Cervantes responsible for the deficit. By some means the money was raised, and the debt was liquidated on the 21st of January 1597. But Cervantes’ position was shaken, and his unbusinesslike habits lent themselves to misinterpretation. On the 6th of September 1597 he was ordered to find sureties that he would present himself at Madrid within twenty days, and there submit to the exchequer vouchers for all official moneys collected by him in Granada and elsewhere. No such sureties being available, he was committed to Seville jail, but was released on the 1st of December on condition that he complied with the original order of the court within thirty days. He was apparently unable to find bail, was dismissed from the public service, and sank into extreme poverty. During a momentary absence from Seville in February 1599, he was again summoned to Madrid by the treasury, but does not appear to have obeyed: it is only too likely that he had not the money to pay for the journey. There is some reason to think that he was imprisoned at Seville in 1602, but nothing positive is known of his existence between 1600 and the 8th of February 1603: at the latter date he seems to have been at Valladolid, to which city Philip III. had removed the court in 1601.
Since the publication of the Galatea in 1585 Cervantes’