The Political character of the Pyrenees corresponds to the Physical character which has been described. The high crest is the bond and division, from the beginning, between two societies which are connected by such common social habits as mountains impose—which therefore fall under similar local customs, which have a common jealousy of the civilized power on the plains below them, and which support each other in a tacit way against the stranger, yet which, from the beginning, have different governments and (especially in the high central part) deal with different corporate traditions—to the north the Béarnese, to the south Aragon. The easier passes to the west and the east of the chain permit a more or less homogeneous community to straddle across either end of the mountains, and to hold upon both slopes the sea roads that pass along the Atlantic and the Mediterranean. The people thus astraddle of the eastern end have come to be called the Catalans. That astraddle of the western, a highly distinct group of men with language, traditions, and physical characteristics wholly their own, has always been known by some title closely resembling their modern name of Basques.

The foundation, therefore, upon which Pyrenean History is built, or (to use another metaphor), the germ from which it has developed and which explains its course, is a tripartite division of the inhabitants, corresponding, as I shall presently show, to the physical features of the chain: an eastern or Catalan, a western or Basque, and a central group whose characteristic it is to subdivide according to the deep valleys into which it is separated, but which falls into two main societies, the one north of the chain which becomes the group of French counties whose typical government is Béarn, the other south of the chain, which assumes at last for its title “The Kingdom of Aragon.”

The first matter to be noticed with regard to this tripartite division is the exactitude of its boundaries. One might imagine that the language, the habits, and the clear characteristics of any group would merge easily into those of its neighbours upon either side. This is not the case. The Basque type—much the most particular—ceases abruptly upon the watershed between the Gave d’Oloron and the Gave d’Aspe to the north of the range, upon the watershed between the Veral and the Esca to the south of it. The Catalans, with a dialect, mind, and dress wholly their own, are found to the north from the sea up to the Col de Puymorens, and everywhere east of the Carlitte mountains; in the Ariège valley and just over these heights, and on the further side of that Col, they are changed. To the south of the range they extend everywhere from the sea to the valley of the Ribagorza. Cross westward from that Catalan valley to the Esera. There, after hours of scrambling, down by the rocks and deserted tarns, you may towards evening find a man; that man will show the slow gestures, the silence, and the elaborate courtesy of Aragon.

The mountain ridges which divide these various peoples are sufficient to mark their boundaries; but they do not suffice to explain why the Catalan, the Basque, the Aragonese, the Béarnais should cease suddenly here or there. True, the high lateral ridges which are so striking a peculiarity of the Pyrenees form barriers with difficulty passed, but these barriers are found just as high and just as precipitous and savage between two valleys of the same speech and nation as between two of different allegiance. Thus the wild jumble of mountains, “the Enchanted Range,” cuts off the Catalans of Esterri from the Catalans of the Ribagorzana. To pass them is something of a feat for anyone not of these hills—for much of the year they are closed to the native inhabitants. Their passage is hardly more of a task or more precipitous than the passage from Aragonese Venasque to Aragonese Bielsa, or from Béarnais Gabas, in the Val d’Ossau, to Béarnais Urdos, in the Val d’Aspe.

An explanation of the unity which rules over each group, Basque, Central and Catalan, can only be given by referring each to the plains at the mouths of the valleys. It is the towns at the entry of these plains that form the markets and rallying places of the mountaineers and that determine their groupings. Oloron is the link between the two Béarnais valleys I have mentioned. Urgel binds Catalan Andorra to Catalan Esterri. Why, however, the groups should lie exactly where they do it is impossible to determine, for no records reach beyond the Romans. All we can say is that the Pic d’Anie, the first high peak eastward from the Mediterranean, forms the boundary stone of the Basques, as it does the chief physical mark dividing the high central ridge from the easier western passes; that the tangle of difficult and impossible peaks just eastward of the Maladetta are the boundary of Catalan south of the range, the similar but less abrupt tangle of the Carlitte, their boundary upon the north. How these nations arose, whence they wandered, whether their differentiation has arisen upon the spot out of an earlier homogeneity or is due to the conflict of invaders—of all this we know nothing.

The place names of the Pyrenees, like those of all Spain, and half Gascony, do indeed afford a curious speculation which arises from the high proportion of names that are certainly Basque, though out of Basque territory. Of this language I shall write later: for my present purpose the point I would desire the reader to note is the sharp contrast which exists between that idiom and the idioms around it. There is no mistaking a Basque word, and yet these are found in all the Pyrenean range and to the north and south of them in a hundred place names, attached to hills, rivers and towns where Basque has been unknown throughout all recorded history. It is even plausibly suggested that the Latin “Vascones,” the French “Gascon” is equivalent to “Basque,” and the late Mr. York Powell, the Regius Professor of History at Oxford, would say in speaking upon this matter that “Gascon was Latin spoken by Basques.” He possessed that type of education, rare or unknown in our universities, which made him capable of individual judgment in departments of living knowledge where his colleagues could but repeat words taught them from a book. This quality reposed upon a wide acquaintance with all matters of European interest. His diverse reading and considerable travel enabled him to balance human evidence in a way hopeless to his less fortunate neighbours in the University, and his conclusion on this important detail of history has always recurred to me when I have examined some new point in the early history of these mountains. There must, however, be set against the general conclusion that the Basques are the remnant of a people once universal from the Garonne to the Pyrenees, and throughout the Iberian Peninsula, the fact that they present a marked physical type utterly distinct from others upon every side. That a race of such a character, vigorous, attached to the soil, in no way nomadic, should have abandoned a large territory is difficult to believe; moreover, there is no case in all the recorded history of Western Europe of one people ousting another, and the process is manifestly physically impossible, save among nomads. Jews or Arabs could propagate and even believe such a theory. To Europeans it is laughable: the peasants and cities of Europe never have been, nor ever can be, largely displaced.

All we know is that these place names exist throughout Spain and all over the Pyrenees, and that the million or so who speak the language whence such names are derived now occupy a tiny corner only of the vast territory over which those names are spread. The rest is guesswork.

Ignorant as we are of the origin of the differentiation between Basque, Béarnais, Catalan and Aragonese, an historical fact quite certain—though no document proves it—is the extreme antiquity of these classes of men. That all Pyrenean history reposes upon their separate existence must be evident to anyone who has watched the commercial manner, the mercantile vivacity, the whole mentality of the Catalan, and has contrasted it with the quiet chivalry of Aragon. Different military fortunes, different economic outlets, and different accidents of central government may possibly account within the historic period for the contrast between the Aragonese and the people of Béarn, Bigorre, or Comminges. No such forces can account for the gulf that cuts off the Catalan and the Basque at either end of the chain from the inhabitants of its high central portion. Infinite time is the maker of states, and two thousand years could never have determined societies so sharply separate. We must regard their constant and immemorial presence in the Pyrenees as the first and enduring principle to guide us in the history of those mountains.

From this fundamental truth, which leads the prehistoric into the historic, one must proceed to another political fact of high importance, which is that while the watershed of the range has but partially separated customs and local thought, and that only in the centre of the range, it has necessarily served as a political boundary whenever a high civilization found it necessary to establish such a strict line. The boundary and the watershed may not exactly coincide—they do not exactly coincide even in the highly organized condition of modern society; but in the two historical periods of strict policy, the Roman and our own, the crest of the range has marked, and marks, an obvious boundary for most of its length. The political distinction between Hispania and Gaul cut the Basque nation into two, following the mountains from Roncesvalles to the Pic d’Anie: it cut the Catalan people into two, following the water parting from the two Nogueras to the Mediterranean. It followed the central chain, indifferent to the similarity or difference between the northern and the southern valleys. To-day the political distinction between Spain and France follows nearly the same line.

The reason of this was, and is, twofold. First, that a clear physical boundary easily definable and of its nature permanent—the crest of a chain, a broad river, or what not—necessarily recommends itself to a bureaucracy in search of simplicity and economy in the work of a great political machine. We see it in the new countries to-day, where the instinct of organized government for easily definable and exact limits takes refuge in establishing parallels of latitude as state boundaries in the absence of marked physical lines. Secondly, in the case of mountains, and especially of mountains as sharp and as boldly set as are the Pyrenees, the fatigue of climbing, the absence of carriageways, made each valley dependent for its connexion with the central government upon some town of the plains, and the authority of a provincial magistrate could not but run, as ran the physical instruments of his rule, up from Huesca northward to Sallent—for instance, or up from Jaca to Canfranc, and so to the summit of the ridge; or up from Oloron southward to Accous, and so to Urdos. As the messengers, writs, powers of each proceeded, the way would become harder, the progress more doubtful. It was obvious and necessary that the boundary of either jurisdiction should lie upon the pass. And though the inhabitants of the northern and the southern valleys might be accustomed to a regular intercourse across the crest, the Roman agents of a distant central government could not but have depended upon cities far removed to the south and to the north of the watershed, as to-day the police of Tardets, let us say, and of Isaba, two towns of one speech, refer respectively through Pau and through Pamplona to Paris and to Madrid.

It is in the interplay of these two jarring political forces, the permanent national seats of Basque, Catalan, etc., and the use of the range as a political or official boundary, that the political character of the Pyrenees resides; and as their history begins with the Romans, to whom we owe the first knowledge of the Pyrenean people and the first use of the Pyrenean boundary, it will be well to consider it under territories divided as the Romans divided them, by the main range, and to follow first the development of the northern slope.

The historical origins of the French Pyrenees are sharply divided in history by that wall which cuts off all that Rome made from all that Rome inherited. Rome made of the barbarians a new world, but before she began that task Rome had inherited everywhere within a march of the Mediterranean a belt of land whose civilization was similar to, always as old as, and sometimes older than, her own. It was a municipal civilization dependent upon the arts and religion proper to a city state. It built, whether temples or ships, as Rome would build them: it was one thing; it is almost one thing to-day; and its bond at Antioch as at Saguntum, at Marseilles as at Athens or Alexandria, was, and is, the universal water of the Mediterranean. To such cities and their territories Rome fell heir. Little proceeded from her to them save first the sense of unity, and later the Faith, and of the whole system, the belt which stretches from Valencia to Genoa, now broadening to the plains of Nemosus (Nîmes), now narrowing to the rocky ledge of the Portus Veneris (Port Vendres), concerns the first evidence of Pyrenean history; for it was from a corner of this belt—between Tarragona and Narbonne—that the advance of civilization inland and along the Chain proceeded.

A century before the four imperial centuries which made our Christian world, a century before Augustus Cæsar, Rome had fully occupied and impressed that soil—to the south Gerona and the Catalan fields, to the north the rich floor which lies under the Canigou and has come to be called the Rousillon. Thence the Roman advance north of the hills proceeded. The chief town of the sea-plain—whose name “Illiberis” is so strongly Basque in form—Rome took for the central municipality of that plain, and made it the capital of the coastal district. This hill and citadel, at which Hannibal had halted a hundred years before, preserved as a bishopric for thirteen hundred years a memory of the Roman order. Constantine formed its diocese, rebuilt it, gave it his mother’s name of Helena. The sea by which it lived has withdrawn from it. It has sunk to be a little country town, “Elne.” Roscino which lay also upon the coast march of Hannibal, has sunk to something smaller still, yet, by some accident, gave the province, in the dark ages, its name of Roussillon which it still retains. These two towns, the fruitful plain about them, the Port of Venus (which is now Port Vendres), formed the municipal structure of this district, the last corner of the great province whose headship lay at Narbonne. Its nominal boundaries included all the vale of the Tet; it extended as far as now extends the Catalan language, and was bounded, as that is bounded, by the great form of the Carlitte and its high lakes and snows. All between that mountain and the sea, all the eastern decline of the range and the slope north of it, was ancient land, and had been ploughed and held and walled by men of the Mediterranean civilization long before Rome inherited it.

With the much longer stretch that runs from the upper Ariège to the Atlantic it was very different. This was of what Rome made, not of what Rome inherited. Before the coming of Roman government it was barbarous, and the many tribes or petty states, whose number various guesses of antiquity record (they were perhaps as numerous in their subdivision as the valleys), stand in three main groups when first the civilization to the east of them began to record their existence: these three were first the Convenæ, south of Toulouse, and all about the upper waters of the Garonne. Next to these came the Auscians, and finally, over the Basque end of the hills towards the ocean, was the seat of the Tarbelli.

The whole point of view of antiquity differed from ours in speaking of such tribes, nor is it easy to pick out from the scraps of observation that have come down to us the kind of information that we want. Sometimes a name survives, sometimes it does not; sometimes we get a hint of a variety of race, most often we lack it. It is the very meagreness and eccentricity of the information upon a barbarous race and custom which affords such opportunities to our dons for those forms of speculation which they love to put forward as dogma, the most absurd example of which, perhaps, is the interpretation and enlargement of Tacitus’ “Germania.” It is therefore exceedingly difficult to know of what kind were these people beyond the old Roman pale. We do not know what language they spoke. We only know that, like other Gallic communities, they centred round fortified places, that their pacification was easy, and that, like everything else in Western Europe, they were of an unchangeable kind.

The whole district between the Garonne and the Pyrenees came to be called, during the first four centuries of our era, “The Nine Peoples.” The Convenæ are early noted to have attached to them upon their right and upon their left, to east and to west, the Consevanni and the Bigerriones. The first of these were (to follow the high authority of Duchesne) organized as early as the first century; what is now St. Lizier was their old capital and later their bishopric, which takes its present name from Glycerius, a saint of the sixth century. They held all those hills of which St. Girons close by is now the centre. The Bigerriones are not heard of until the mention of them in the Notitia of the fourth century. They must have held Bigorre, and the three valleys which I have called the valleys of Tarbes. Tarbes—then Turba—was their capital, and was and is their bishopric.

The Auscians do not concern us. They and the three groups into which they are later distinguished held the western plains and foot hills. The Tarbelli held both the foot hills and the mountains of the west; their capital was at Dax. They also split into, or are later recognized as three separate groups, making up with the two other sets of three “The Nine Peoples,” under which title all this country below the Pyrenees became permanently known. But of the three only the Civitas Benarnensium, whence we get the name Béarn, and the Civitas Elloronensium, with its capital at Iloro, which has become Oloron, concern us. The capital and soon the bishopric of the Civitas Benarnensium was at Lescar, as far as we can make out, and Lescar bore the chief sanctity in Béarn until that country was swept by the Reformation. The sovereigns of Béarn were buried there, even the Protestant sovereigns, and it remained a bishopric, whose bishop was the President of the Parliament of Béarn, until the Revolution; but it was the Reformation which destroyed its original character of a capital.

We have, therefore with the earliest ages of our civilization, five peoples holding the northern Pyrenees, the Consevanni, the Convenæ, the people of Bigorre, the Béarnese, and the Elloronians.

It is remarkable that in such a list, our Roman originators and their geographers overlooked the Basques. The category ends precisely at the present limit of the Basque tongue. For the Val d’Aspe, of which Oloron is the town, is the first French-speaking valley. Why it is that we hear nothing of the Basques it is difficult to say, especially as the second of the great Roman military roads went right through their country. Bayonne, which is the Basque’s town of the plains on the north, is heard of in the fifth century. It has a garrison; but no bishopric until the tenth. Pamplona, which is their town on the south, was known before the beginnings of our Christian history. But the Basques themselves are not known to us from the Romans. The name of the Consevanni survives locally. The country round St. Girons is still all one country-side and called the “Conserans.” Of the Convenæ we have a pleasant legend in St. Jerome telling how Pompey got together all the brigands of the mountains, drove them northward hither, and forced them into a garrison (a stronghold which, like Lyons and the rest, was one of the many “Lugdunums”). It was destroyed early in the dark ages, and later revived by St. Bertrand, a little way off in his Episcopal town. Their name survives in the district of Comminges. The Béarnese name of course survives and so does the Bigorrean, while the Elloronean, though no longer the title of a district, is preserved in the town name of Oloron.

All this country, not only that of the five tribes along the mountains, but the whole territory occupied by the nine peoples (who afterwards became twelve), lay in a profound peace under Roman rule, and we may be certain of its increasing wealth throughout the first four great formative centuries of our era.

The advance of Rome upon the Spanish side was of a very different kind. Rome, after the Carthaginian wars, inherited broad belts of civilized and half civilized land. All the Mediterranean slope below the mouth of the Ebro, and a belt quite three days’ marches wide inland to the north of that river, was full of ancient populated towns, alive with the full civilization common to every shore of the inland sea. So, we may be certain, were the broad plains of the south where the most complete and earliest absorption of the Celt-Iberian in Roman speech and ideas took place. The advance into the north-west and therefore along the Pyrenees covered more than a century of strict and perpetual warfare, which was intermixed by the civil wars of the Roman commanders. The extremities of the Asturias were reached in the century before the birth of our Lord, but the advance was not, as upon the north, a rapid expansion beyond the old boundaries. It took the form of siege after siege and battle after battle, in which those numerous and crushing defeats, which Rome (like every truly military power) reckoned to be a necessary part of history, interrupted the slow progress of her law. The Celt-Iberian towns were walled and strong; their resistance was painful and tenacious; there was no sudden illumination of a willing people by a new culture, such as had taken place in Gaul, rather was northern Spain kneaded by generations of warfare into the stuff of the Empire.

When the work was accomplished, it was complete throughout the Peninsula; and though the silent strength of the Basques prevented the Roman language from invading their valleys, the administration of the whole territory south of the Pyrenees must have been as exact and as bureaucratic as that to the north of it. There was, however, this great difference due to local topography between the Spanish and the French hills, that the municipalities upon which Rome stretched her power, as upon pegs, were less common, were farther apart, and approached less nearly to the central ridge upon the southern than upon the northern side. What you see to-day south of the Pyrenees is what you might have seen at any time in the last 2000 years—a very few scattered towns, still the centres of government, and all the rest rare isolated villages living their own life, free from the criminal, and, by a regular payment of small taxes, half independent of the civil law. Alone of the true mountain sites, Jaca in the middle, Pamplona and Urgel at either extremity, were bishoprics. Huesca, St. Laurence’s town, a fourth centre, is in the plains. For the rest the confused storm of hills ending in those parallel ranges, pushed right out on to the burnt flats of the Ebro, forbade the establishment of a municipal civilization.

Upon all this land to the north and to the south of the mountains came, after five hundred years of a high civilization, the slow decline of culture, and the infiltration of the barbarians. In a sense the nominal divisions between the barbaric kingdoms has its importance, for they help us to understand changes of dynasty and of custom. But they were of no political effect. The mass of the people knew little of the chance soldiers who, with their mixed retinues of Roman, Breton, German, Slav, and the rest—some able, some not able to read the letters of Rome—sat in the old seats of office, issued their writs through the still surviving Roman Bureaucracy and from palaces which were but those of decayed Roman governors.

For the greater part of Western Europe, and especially of Gaul, this process of decay was one into which Europe slowly dipped as into a bath of sleep, and out of which it rose more rapidly through the energy of the Crusades and of the renewed Pontificate into the splendour of the Middle Ages. But the Pyrenees suffered in this matter a peculiar fate. When Spain was overrun by the Mohammedan, and when in the first generation of the eighth century the Asiatic with his alien creed and morals had even swept for a moment into Gaul, the Pyrenees became a march: at first the rampart, later, when they were fully held, the bastion of our civilization against its chief peril. It is this episode by which the Pyrenees became the military base of the advance against Islam—an episode covering the whole life of Charlemagne and after him the ninth, tenth, and eleventh centuries—which gives them their legendary atmosphere and fills all their names with romance.

The northern slope, during the long business by which Gaul became itself again, was but a remote border province. The new life of Gaul, after the shock which had so nearly destroyed Europe was over, sprang from Paris. The influence of Paris radiating upon every side built up again accuracy of knowledge, unity of government, and general law. To this influence the Pyrenees seemed the most remote of boundaries. The valleys were little affected by the growth of the French Monarchy; they remained for centuries broken into a maze of half-republican customs, of tiny independent lordships, guarded and menaced by separate and jealous walled municipalities upon the plains—all of this vaguely and slowly coalesced into larger districts, doubtful of their sovereignty and perpetually struggling upon their boundaries and their sub-boundaries.

In this development nothing was more striking than the way in which this remote border at first looked rather to the south, where the interest of religious war was ever present, than to the north, whence government was slowly coming towards it. The French Pyrenees fought and felt with the whole range, not with the plains. Jaca in the worst time, when it was the only mountain bishopric free from the Mohammedan, counselled with and was perhaps suffragan to Eauze. Urgel sat in the provincial Synod of Narbonne. As the success of the Reconquista pushed the noise of crusade further and further from the range, the northern valleys looked more and more towards their northern towns. Their nominal allegiances grew stricter—as of Foix to Toulouse—and every French bishopric was bound more and more to its northern metropolitan, the Spanish sees to the new metropolitans of the Ebro.

At last an issue was joined between Northern and Southern France of the first moment to the unity of Gaul itself and of all Christendom. The Crusades, the knowledge of the East, the awakening of the intelligence and of its appetites, had bred throughout the wealthy towns of what had been from the beginning Roman land, a desire to be rid of the restraints of fixed religion. The South of France began to move towards its pagan past. It was a movement which had already had its strange echo in the north, a movement which in England had only been pulled up at the last moment by the martyrdom of St. Thomas. Here in Gaul, in the sunlight, and backed by so much gold, the rational and sensual revolt became a larger thing, and when various sources of disruption, speculation, and achievement had met in one stream, it was commonly called the Albigensian movement. The issue was decided, after heavy fighting, in the early thirteenth century, and the victory was with the cause of the unity of Europe. Toulouse (the true centre of the storm) and its lord were conquered. Northern barons swept down, held no small part of the southern land, and from that time onward the French Pyrenees are normally dependent on Paris.

Two exceptions survived, the straddle of the Basque across the chain and the straddle of the Catalan. The Basque had his country of Navarre upon either side of the chain; with it went Béarn, and these were independent of the French crown. The Catalan, able to traverse the chain by the flat floor of the Cerdagne, preserved the unity of his mountain province, and the Roussillon counted with Spain. Apart from this easy passage into the Roussillon from the south, by way of the Cerdagne, the isolation of the Roussillon was the more easily accomplished from the long spur of the Corbieres, which runs north and east towards the sea and cuts off from France the wealthy plain of which Elne had once been, Perpignan had later become, the capital.

This arrangement endured, in name at least, until the seventeenth century. The last heir of Navarre was also the heir nearest to the French throne at the close of the religious wars, and as Henry IV of France he united the two crowns. A man who, as a boy, might have rejoiced at that union, could have lived to see, under Mazarin, the signing of the Treaty of the Pyrenees, which gave the Roussillon also to the French Crown. The date of this final arrangement coincided with what is ironically called “The Restoration” in England: this date, which definitely closed the power of the English Monarchy, and substituted for it the power of a wealthy oligarchic class, coincides throughout Europe with the struggle between absolute central government for the equal service of all, and local aristocratic custom. In England the latter conquered; in Spain and France the debate was decided in favour of the former.

Such centralized governments could but further define and insist upon a new boundary, and from that time onward, for 250 years that is, the Pyrenees have been once more as they were under the clear administration of Rome, a fixed political boundary; and, save certain exceptions that will be mentioned later, everything north of the chain has been administered from Paris, and has slowly accepted, side by side with the local tongue, the tongue of Northern France and the habit of a centralized French government.

South of the chain the process by which Christendom recrystallized out of the flux of the dark ages, followed a different course; it was a process to which Spain owes all her national characteristics, for out of the mountains a Spanish nation was formed, and from its various communities, as from roots, the Catholic kingships grew southward until they once more reached Gibraltar.

To understand this process, it is necessary to consider factors absent in the topography of the Gaulish plains, and especially the factor of that unconquerable tangle of mountains which occupies all the north-western triangle of the Peninsula.

The ocean boundaries of the Iberian quadrilateral are nearly square to the points of the compass. It is not so with the internal divisions that mark off its central part. Here the edges of the high and arid plateaux, the deep trenches of the rivers, the mountain ranges, the boundaries of the plains at their feet, run slantways from north-east to the south-west. This slant determined the boundaries of Mohammedan expansion, while the Asiatics and Africans still retained the energy to advance; it determined the successive frontiers of the Reconquest, as our race slowly ousted the invader and reached at last the sea-coast of Granada. The Arab and the Moor were masters of Narbonne and all the Roussillon on the east, when, on the west, they could not cross the mouth of the Mulio a hundred miles to the south. They were at Jaca within a day’s march of the watershed along the Roman Road, when, to the immediate west of it, they could not hold Fuente; they could not even reach Pamplona, though that western town is two marches at least from the main crest. Toledo was reconquered a generation before Saragossa, though Saragossa is by nearly two degrees more northerly, because Toledo was west of Saragossa. The last Mohammedan kingdom was crowded, after the thirteenth century, into the extreme south-east, as the surviving remnant of the free Europeans of the Peninsula had been crowded into the extreme north-west in the eighth.

If the boundaries of undisputed Mohammedan rule be traced for various dates, the receding wave will be found in general to follow curves that lead, like the main features of the land, from the north-east downwards towards the Atlantic.

This main character in the geography and history of Spain, the south-westerly trend of the mountain ridges, largely determined the fortunes of those fighting bands of mountaineers who ceaselessly pressed southward until they had wholly driven out the invader and reconstituted the unity of Europe. It determined the first advance to be, not from the Pyrenees, but from the Asturias, and the first captain connected with the Christian resistance after the overwhelming of all that civilization, Pelayo (from whose blood Leon, Castille, Aragon, and Navarre descend) had his stronghold, not in the Pyrenees, but a week’s march to the west, along the Biscayan coast at Cangas. Within the decade of the invasion he had checked the invader in his own hills at Covadonga.

All the eighth century is full of that successful spirit in the north-west—but nowhere else. Alfonso, the husband of Pelayo’s daughter, struck the note with his boast, “No pact with the infidel,” and the tradition or prophecy that Christendom would regain the south, springs from him. He conquered down to the Douro, over what is to-day the mountain frontier of Portugal; he began those long cavalry raids into the heart of Moorish land. He rode into Astorga, into Zamora, into Segovia itself—within sight of the central range of the Guadarama: riding back with booty, harassed and harassing, nowhere permanently fixing himself save in the towns of the west, upon the Lower Douro, but building on the ridges of his defence, those block-houses, the “Castille” from which, long after, the frontier province began to take its name.

All the ninth century that spirit grew. The body of St. James was found under the Star at Compostella—its shrine became the national sacrament as it were, a perpetual refreshment for arms, and a symbol, in its wild isolation among the rocks of Galicia, of the impregnable places from which the Reconquista drew its ardour. The advance continued. The frontier counties consolidated and were named.

Leon was permanently held, Burgos was founded. If one takes for a date the opening years of the tenth century, just after Alfred had saved England also from the pagan, and just after the Counts of Paris had saved northern Gaul, there is a full Spanish kingdom standing up against the Mohammedan power, a king has been crowned in Leon and has died in peace at Zamora. The cavalry raids have pushed—once at least—to Toledo. All the north-west lay permanently Christian beyond a line that ran from the corner of Gaul to the Douro and down the Douro to the sea; and this united triangle of Roman land formed a base from which the pushing back of the alien could proceed.

How did this disposition of forces affect the Pyrenees? Let it first be noted that the newly organized Christian country lay wholly to the west of the range. In the Pyrenees themselves the Mohammedan flood had washed every valley. The crest had been traversed and retraversed; both slopes were for a moment held by the invaders. Abd-ur-Rahman had sent or led his thousands by the Roman roads of Roncesvalles and Urdos and over the Ostondo and the lower passes of the west. The mule tracks of these rocks had been twice crowded with the white cloaks of the Arabs. In the east, Narbonne was held for fifty years, and with it all the Catalans. Even in the high centre of the chain, where there is no passing between the Somport and the Cerdagne, wherever there was something to rob or to destroy, the invaders had penetrated. There was not here, as in the Asturias, untouched land.

When the crest of the wave retreated, when the Mohammedan came back defeated from Gaul, the high valleys attained—it may be guessed—a savage independence.

Jaca has legends of its battle at the very beginning of the Independence, before Charlemagne had come to the rescue, and from all the valleys of the Sobrarbe, bands of men must have been perpetually volunteering for skirmishes down into the plains. Navarre was the natural leader of the movement, the largest and the most fertile belt of Christian land, but the little lordship upon the Aragon, fighting down south and east towards the Ebro, the western count of the Asturias fighting down south and west, cut off the advance of the Basques; and though Navarre in the period of birth and turmoil which is that of Gregory VII’s reform of the Papacy, of the establishment in England and in Sicily of Norman power, and of preparation for the Crusade, was the head of all the southern Pyrenees and called itself an “Empire,” it was blocked by the double line of advance, and the Basques, upon the foundation of whose tenacity and courage, as upon a pivot, the Reconquest had proceeded, took little more part in the wars; but the Basque strip of Navarre gave its first king to Aragon, and the son of that first king, Sancho, raided so far as Huesca and was killed beneath its walls; his son again, Peter, took the town two years later just as the hosts of Europe were gathering for that first great march upon Jerusalem which threw open the curtain of the Middle Ages; and his son, Alfonso (who had united in one crown Leon and Aragon), went forward under his great name of the Batallador, and twenty years later (1119) swept into Saragossa, the last of the Mohammedan strongholds in the north.

Thus were the west and the centre of the Spanish slope recovered for our race and civilization.

Meanwhile Catalonia upon the east had been since Charlemagne, since the early ninth century, a march of Christendom; but it was not until the same creative period which had brought forth the leadership of Navarre and the advance from Aragon, the Normans, Hildebrand, and the resurrection of Europe, that Catalonia began to go forward. Its first true monarch was Berenguer the Old, who lived round and about the date of Hastings, and was first master of the whole province. He also founded and maintained the Cortes of Barcelona. His son, for a moment, raided the Balearics, and when he died Catalonia and Aragon, united under one crown, saw the alien finally driven from these mountains. All the plain from far beyond the base of the hills was now permanently held by strong and united kingships which pressed forward to the Ebro valley, and finally saved all the Spanish province of Europe. A lifetime later, the last of the foreign armies had been broken at Navas de Tolosa. Far off in the south Islam lingered, tolerated and on sufferance, but Spain was reconquered. For just 500 years Spain, a quarter of all that makes up our civilization, had lain in peril between our religion and the other.

I have said that with the thirteenth century, the Albigensian crusade upon the north, the destruction of Islam upon the south (the two successes were contemporaneous), the Pyrenees ceased for ever to be a march between two civilizations, and became a mere political boundary between two provinces of Europe; and I have said that the nature of that boundary was finally fixed in the seventeenth century, or rather during a period which stretches from the close of the sixteenth to just after the middle of the seventeenth.

If that political boundary be examined to-day it will be found to coincide with the watershed, save at certain particular points, the character of which merits examination.

I take for my boundaries, as throughout this book, Mount Urtioga on the west, and the beginning of the Alberes on the east, which may conveniently be placed at the Couloum.

In this distance, there is a slight discrepancy between the political boundary and the watershed here and there in the Basque valleys. Mount Urtioga itself, though upon the watershed, is entirely in Spain, and the sources of the torrents which feed the valley of Baigorry all rise a mile or so beyond the political frontier, which is here composed of two straight conventional lines.

The head waters of the Nive are wholly in Spain also, as is all the left bank of that river to a point four miles below Val Carlos. The right bank, however, is French, so far as the torrent Garratono. Thence forward from the sources of that torrent (that is, upon the Atheta) the frontier now follows the watershed, now leaves the very head-springs of the torrents in Spain until, a few miles further east, it makes a considerable invasion of Spanish territory, not because the frontier itself bends, but because the watershed here goes northward in a half circle. All the upper valley of the Iraty is politically in France; but from the Pic d’Orhy, where a definite ridge begins, it follows the frontier strictly for mile after mile (with the exception of a curious little enclave which gives Spain two or three hundred yards of the head-waters of the Aspe), and there is no further exception throughout all the high Pyrenees until one strikes the curious anomaly of the Val d’Aran.

I have said in describing the physical structure of the Pyrenees that the two main axes of those mountains were joined by a sort of fault, a serpentine bridge of high land which united them from the Sabouredo to a point ten miles northward, the Pic de l’Homme overhanging the Pass of Bonaigo. The valley caught on the French side of this twist is the Val d’Aran, containing the upper waters of the French river Garonne. Geographically of course it is French, but politically it is Spanish so far as a certain gorge where is a bridge called the King’s Bridge, and where the Garonne pours through a narrow gate of rock into its lower valley. The story goes that when the Treaty of the Pyrenees was in act of negotiation someone said diplomatically and casually to the French negotiators, “The Val d’Aran of course you regard as Spanish,” and they, knowing no more of these mountains than of the mountains of the moon, said, “Of course.” The true reason is rather that the gate in the mountains cuts off this upper valley from the lower gorges of the river much more than the low, easy, and grassy saddle of the Bonaigo cuts it off from the Spanish valley of Eneou just to the east of it: and though the Val d’Aran may be geographically or rather hydrographically French, it is topographically Spanish, which is as though one were to say that Almighty God made it so.

Another exception and a big one to the rule that the frontier follows the watershed, is, of course, to be found in the French Cerdagne. The true watershed here is coincident with the frontier as far as the Pic de la Cabanette in latitude 42° 35′ 30″. The watershed then goes on over the Port de Saldeu, along the crest of the Port d’Embalire to the Pic Nègre, and there it turns to the east along the ridge across the saddle of which goes the high road over the Col called Puymorens. It follows that ridge, not to the summit of the Carlitte, but to a lower peak called the Madides, three miles to the north-east, runs along two miles of a high rocky ledge to the Pic de la Madge and then there follows a difficult sort of hydrographical No Man’s Land, the centre of which is the great marsh of Pouillouse, nor can you tell exactly where the watershed is for some miles in the forest below that marsh, for the same damp flat ground sends water into the valley of the Tet and into the valley of the Segre. Three miles to the south-west, however, it is clearly defined again in a low rounded lump of wooded land, it passes over the flat Col de la Perche and then follows the crest still going south-west up to the Pic d’Eyne, where again it becomes the frontier, and the frontier it remains until it reaches the Mediterranean.

From the Pic de la Cabanette, all the way to the Pic d’Eyne, France and Mazarin politically took in by the Treaty of the Pyrenees a belt to the south of the watershed and extending down to a conventional line which left Bourg-Madame French and Puigcerdá Spanish; an exception in this is a small strip beyond the Pic de la Cabanette, on the left bank of the Ariège, which, though geographically French, was given to Andorra, so that Andorra might smuggle more comfortably over the passes.

The causes of this annexation of the French Cerdagne by Mazarin are clear enough when one remembers that the Roussillon (which is geographically French) passed to France by the same treaty. There is no way from the valley of the Ariège into the Roussillon except by going round this corner of the Cerdagne, at least no practicable carriage way; the only other way is the difficult and high short cut described later in this book. If the frontier be carefully noted, it will be seen that it is designed merely to preserve a right of passage over this road. Jurisdiction was only claimed by France over the villages, and Llivia, being a town, stands in an island of Spanish territory in the midst of the French Cerdagne, as will be seen later when I speak of this district in detail.

Such is the present political aspect of the Pyrenees, with Toulouse for their great French town in the plains, 60 miles away to the north, Saragossa for their great Spanish town in the plains, 100 miles away to the south, a string of towns just at their feet (Bayonne, Pau, Tarbes, St. Girons, etc.) on the northern side; on the south a rarer and less connected group (Pamplona, Huesca, Barbastro, Lerida, etc.); and against the Mediterranean the district of Gerona, shut in by the Sierra del Cadi (with its outposts) and the Alberes upon the Spanish side, the town of Gerona its capital; the Roussillon, with Perpignan for its capital, shut in between the Alberes and the Corbieres on the French side.