A. 5.--For the a.s.signment of the name Ermine Street to the Great North Road there is no ancient authority.[196] All we can say is that this theory is more probable than that set forth by Geoffrey of Monmouth.

That the road existed in Roman times is certain, as London and York were the two chief towns in the island; and direct communication between them must have been of the first importance, both for military and economical reasons. Indeed it is probably older yet. (See p. 117.) But, with the exceptions already pointed out, the nomenclature of the Romano-British roads is almost wholly guess-work. Some archaeological maps show additional Watling Streets and Ermine Streets branching in all directions over the land,[197] presumably on the authority of local tradition. And these traditions may be not wholly unfounded; for the same motives which made the English immigrants of one district ascribe the handiwork of by-gone days to mythological powers might operate to the like end in another.

A. 6.--The origin of the names Ryknield Street and Akeman Street is beyond discovery;[198] but that of the Icknield Street is almost undoubtedly due to its connection with the great Icenian tribe, to whose territory it formed the only outlet.[199] By them, in the days of their greatness, it was probably driven to the Thames, the more southerly extension being perhaps later. It was never, as its present condition abundantly testifies, made into a regular Roman "Street."

The final syllable may possibly, as Guest suggests, be the A.S. _hild_ = war.

A. 7.--Besides these main routes, a whole network of minor roads must have connected the mult.i.tudinous villages and towns of Roman Britain, a fact which is borne witness to by the very roundabout route often given in the 'Itinerary' of Antoninus between places which we know were directly connected.[200] Moreover this network must have been at least as close as that of our present railways, and probably approximated to that of our present roads.

SECTION B.

Romano-British towns--Ancient lists--Methods of identification--Dense rural population--Remains in Cam valley--Coins--Thimbles--Horseshoes.

B. 1.--Of these many Romano-British towns we have five contemporary lists; those of Ptolemy in the 2nd century, of the Antonine 'Itinerary' in the 3rd, of the 'Not.i.tia'[201] in the 5th, and those of Nennius and of the Ravenna Geographer, composed while the memory of the Roman occupation was still fresh. Ptolemy and Nennius profess to give complete catalogues; the 'Itinerary' and 'Not.i.tia' contain only incidental references; while the Ravenna list, though far the most copious, is expressly stated to be composed only of selected names. Of these it has no fewer than 236, while the 'Not.i.tia' gives 118, Ptolemy 60, and Nennius 28 (to which Marcus Anch.o.r.eta adds 5 more).

B. 2.--With this ma.s.s of material[202] it might seem to be an easy task to locate every Roman site in Britain; especially as Ptolemy gives the lat.i.tude (and sometimes the longitude[203] also) of every place he mentions, and the 'Itinerary' the distances between its stations. Unfortunately it is quite otherwise; and of the whole number barely fifty can be at all certainly identified, while more than half cannot even be guessed at with anything like reasonable probability.

To begin with, the text of every one of these authorities is corrupt to a degree incredible; in Ptolemy we find _Nalkua_, for example, where the 'Itinerary' and Ravenna lists give _Calleva_; _Simeni_ figures for _Iceni_, _Imensa_ for _Tamesis_. The 'Itinerary' itself reads indiscriminately _Segeloco_ and _Ageloco_, _Lagecio_ and _Legeolio_; and examples might be multiplied indefinitely. In Nennius, particularly, the names are so disguised that, with two or three exceptions, their identification is the merest guess-work; _Lunden_ is unmistakable, and _Ebroauc_ is obviously York; but who shall say what places lie hid under _Meguaid_, _Urnath_, _Guasmoric_, and _Celemon_?

And if this corruption is bad amongst the names, it absolutely runs riot amongst the numbers, both in Ptolemy and the 'Itinerary,' so that the degrees of the former and the distances of the latter are alike grievously untrustworthy guides. Ptolemy, for example, says that the longest day in London is 18 hours, an obvious mistake for 17, as the context clearly shows. There is further the actual equation of error in each authority: Ptolemy, for all his care, has confused Exeter (_Isca d.a.m.noniorum_) with the more famous _Isca Silurum_ (Caerleon-on-Usk); and there are blunders in his lat.i.tude and longitude which cannot wholly be ascribed to textual corruption. Still another difficulty is that then, as now, towns quite remote from each other bore the same name, or names very similar. Not only were two called _Isca_, but three were _Venta_, two _Calleva_, two _Segontium_, and no fewer than seven _Magna_; while _Durobrivae_ is only too like to _Durocobrivae_, _Margiodunum_ to _Moridunum_, _Durnovaria_ to _Durovernum_, etc. The last name even gets confounded with _Dubris_ by transcribers.

B. 3.--In all the lists we are struck by the extraordinary preponderance of northern names. Half the sites given by Ptolemy lie north of the Humber, and this is also the case with the Ravenna list, while in the 'Not.i.tia' the proportion is far greater. In the last case this is due to the fact that the military garrisons, with which the catalogue is concerned, were mainly quartered in the north, and a like explanation probably holds good for the earlier and later lists also. Nennius, as is to be expected, draws most of his names from the districts which the Saxons had not yet reached; all being given with the Celtic prefix _Caer_ (=city).

B. 4.--Amid all these snares the most certain identification of a Roman site is furnished by the discovery of inscriptions relating to the special troops with which the name is a.s.sociated in historical doc.u.ments. When, for example, we find in the Roman station at Birdoswald, on the Wall of Hadrian, an inscription recording the occupation of the spot by a Dacian cohort, and read in the 'Not.i.tia'

that such a cohort was posted at _Amboglanna per lineam Valli_, we are sure that Amboglanna and Birdoswald are identical. This method, unfortunately, helps us very little except on the Wall, for the legionary inscriptions elsewhere are found in many places with which history does not particularly a.s.sociate the individual legions thus commemorated.[204] However, the special number of such traces of the Second Legion at Caerleon, the Twentieth at Chester, and the Sixth at York, would alone justify us in certainly determining those places to be the Isca, Deva, and Eborac.u.m given as their respective head-quarters in our doc.u.mentary and historical evidence.

B. 5.--In the case of York another proof is available; for the name, different as it sounds, can be traced, by a continuous stream of linguistic development, through the Old English Eorfowic to the Roman _Eborac.u.m_. In the same way the name of _Dubris_ has unmistakably survived in Dover, _Lemannae_ in Lympne, _Regulbium_ in Reculver.

_Colonia, Glevum_, _Venta, Corinium, Danum_, and _Mancunium_, with the suffix "chester,"[205] have become Colchester, Gloucester, Winchester, Cirencester, Doncaster, and Manchester. Lincoln is _Lindum Colonia_, Richborough, _Ritupis_; while the phonetic value of the word London has remained absolutely unaltered from the very first, and varies but slightly even in its historical orthography.

B. 6.--With names of this cla.s.s, of which there are about thirty, for a starting-point, we can next, by the aid of our various lists (especially Ptolemy's, which gives the tribe in which each town lies, and the 'Itinerary'), a.s.sign, with a very high degree of probability, some thirty more--similarity of name being still more or less of a guide. For example, when midway between _Venta_ (Winchester) and _Sorbiodunum_ (Sarum) the 'Itinerary' places _Brige_, and the name _Broughton_ now occupies this midway spot, _Brige_ and _Broughton_ may be safely a.s.sumed to be the same. This method shows Leicester to be the Roman _Ratae_, Carlisle to be _Luguvallum_, Newcastle _Pons Aelii_, etc., with so much probability that none of these identifications have been seriously disputed amongst antiquaries; while few are found to deny that Cambridge represents _Camboric.u.m_,[206] Huntingdon (or G.o.dmanchester) _Durolipons_, Silchester _Calleva_, etc. A list of all the sites which may be said to be fairly certified will be found at the end of this chapter.

B. 7.--Beyond them we come to about as many more names in our ancient catalogues of which all we can say is that we know the district to which they belong, and may safely apply them to one or other of the existing Roman sites in that district; the particular application being disputed with all the heat of the _odium archaeologic.u.m_. Thus _Bremetonac.u.m_ was certainly in Lancashire; but whether it is now Lancaster, or Overborough, or Ribchester, we will not say; _Caesaromagum_ was certainly in Ess.e.x; but was it Burghstead, Widford, or Chelmsford? And was the original _Camalodunum_ at Colchester, Lexden, or Maldon?

B. 8.--And, yet further, we find, especially in the Ravenna list, mult.i.tudes of names with nothing whatever to tell us of their whereabouts; though nearly all have been seized upon by rival antiquaries, and ascribed to this, that, and the other of the endless Roman sites which meet us all over the country.[207]

B. 9.--For it must be remembered that there are very few old towns in England where Roman remains have not been found, often in profusion; and even amongst the villages such finds are exceedingly common wherever excavations on any large scale have been undertaken. Thus in the Cam valley, where the "coprolite" digging[208] resulted in the systematic turning over of a considerable area, their number is astounding, proving the existence of a teeming population. Many thousands of coins were turned up, scarcely ever in hordes, but scattered singly all over the land, testifying to the amount of petty traffic which must have gone on generation after generation. For these coins are very rarely of gold or silver, and amongst them are found the issues of every Roman Emperor from Augustus to Valentinian III.

And, besides the coins, the soil was found to teem with fragments of Roman pottery; while the many "ashpits" discovered--as many as thirty in a single not very large field--have furnished other articles of domestic use, such as thimbles.[209] Even horseshoes have been found, though their use only came in with the 5th century of our era.[210]

B. 10.--Now there is no reason for supposing that the Cam valley was in any way an exceptionally prosperous or populous district in the Roman period. It contained but one Roman town of even third-cla.s.s importance, Cambridge, and very few of the "villas" in which the great landed proprietors resided. The wealth of remains which it has furnished is merely a by-product of the "coprolite" digging, and it is probable that equally systematic digging would have like results in almost any alluvial district in the island. We may therefore regard it as fairly established that these districts were as thickly peopled under the Romans as at any other period of history, and that the agricultural population of our island has never been larger than in the 3rd and 4th centuries, till its great development in the 19th.

SECTION C.

Fortification of towns late--Chief Roman centres--London--York--Chester--Bath--Silchester--Remains there found--Romano-British handicrafts--Pottery--Basket work--Mining--Rural life--Villas--Forests--Hunting dogs--Husbandry--Britain under the _Pax Romana_.

C. 1.--The profound peace which reigned in these rural districts is shown by the fact that Roman weapons are the rarest of all finds, far less common than the earlier British or the ensuing Saxon.[211] At the same time it is worthy of note that every Roman town which has been excavated has been found to be fortified, often on a most formidable scale. Thus at London there still remains visible a sufficiently large fragment of the wall to show that it must have been at least thirty feet high, while that of Silchester was nine feet thick, with a fosse of no less than thirty yards in width. And at Cirencester the river Churn or Corin (from which the town took its name _Corinium_) was made to flow round the ramparts, which consisted first of an outer facing of stone, then of a core of concrete, and finally an earthen embankment within, the whole reaching a width of at least four yards.

It is probable, however, that these defences, like those of so many of the Gallic cities, and like the Aurelian walls of Rome itself; belong to the decadent period of Roman power, and did not exist (except in the northern garrisons and the great legionary stations, York, Chester, and Caerleon) during the golden age of Roman Britain.[212]

C. 2.--Their circuit, where it has been traced, furnishes a rough gauge of the comparative importance of the Roman towns of Britain.

Far at the head stands London, where the names of Ludgate, Newgate, Aldersgate, Moorgate, Bishopsgate, and Aldgate still mark the ancient boundary line, five miles in extent (including the river-front), nearly twice that of any other town.[213] And abundant traces of the existence of a flourishing suburb have been discovered on the southern bank of the river. To London ran nearly all the chief Roman roads, and the shapeless block now called London Stone was once the _Milliarium_ from which the distances were reckoned along their course throughout the land.[214]

C. 3.--The many relics of the Roman occupation to be seen in the Museum at the Guildhall bear further testimony to the commercial importance of the City in those early days, an importance primarily due, as we have already seen, to the natural facilities for crossing the Thames at London Bridge.[215] The greatness of Roman London seems, however, to have been purely commercial. We do not even know that it was the seat of government for its own division of Britain. It was not a Colony, nor (in spite of the exceptional strength of the site, surrounded, as it was, by natural moats)[216] does it ever appear as of military importance till the campaign of Theodosius at the very end of the chapter.[217] In the 'Not.i.tia' it figures as the head-quarters of the Imperial Treasury, and about the same date we learn that the name Augusta had been bestowed upon the town, as on Caerleon and on so many others throughout the Empire, though the older "London" still remained unforgotten.[218]

C. 4.--But, so far as Britain had a recognized capital at all, York and not London best deserved that name. For here was the chief military nerve-centre of the land, the head-quarters of the Army, where the Commander-in-Chief found himself in ready touch with the thick array of garrisons holding every strategic point along the various routes by which any invader who succeeded in forcing the Wall would penetrate into the land. At York, accordingly, the Emperors who visited Britain mostly held their court; beginning with Hadrian, who here established the Sixth Legion which he had brought over with him, possibly incorporating with it the remains of the Ninth, traces of which are here found. And here it remained permanently quartered to the very end of the Roman occupation, as abundant inscriptions, etc. testify. One of these, found in the excavations for the railway station, is a bra.s.s tablet with a dedication (in Greek) to _The G.o.ds of the Head Praetorium_ [[Greek: theois tois tou haegemonikou praitoriou]], bearing witness to the essential militarism of the city.

C. 5.--A Praetorium, moreover, was not merely a military centre. It was also, as at Jerusalem, a Judgment Hall; and here, probably, the _Juridicus Britanniae_[219] exercised his functions, which would seem to have been something resembling those of a Lord Chief Justice.

Precedents laid down by his Court are quoted as still in force even by the Codex of Justinian (555). One of these incidentally lets us know that the Romans kept up not only a British Army, but a British Fleet in being.[220] The latter, probably, as well as the former, had its head-quarters at York, where the Ouse of old furnished a far more available waterway than now. Even so late as 1066 the great fleet of Harold Hardrada could anchor only a few miles off, at Riccall: and there is good evidence that in the Roman day the river formed an extensive "broad" under the walls of York itself. As at Portsmouth and Plymouth to-day, the presence of officers and seamen of the Imperial Navy must have added to the military bustle in the streets of Eborac.u.m; while tesselated pavements, unknown in the ruder fortresses of the Wall, testify to the softer side of social life in a garrison town.

C. 6.--Chester [Deva] was also a garrison town, the head-quarters of the Twentieth Legion; so was Caerleon-upon-Usk [Isca], with the Second. A detachment was almost certainly detailed from one or other of these to hold Wroxeter [Uriconium], midway between them;[221] thus securing the line of the Marches between the wild districts of Wales and the more fertile and settled regions eastward. And the name of Leicester records the fact (not otherwise known to us) that here too was a military centre; probably sufficient to police the rest of the island.[222]

C. 7.--Gloucester, Colchester, and Lincoln, as being Colonies, may have been also, perhaps, always fortified, and possibly garrisoned.

But in the ordinary Romano-British town, such as London, Silchester, or Bath,[223] the life was probably wholly civilian.

The fortifications, if the place ever had any, were left to decay or removed, the soldiery were withdrawn or converted into a mere _gendarmerie_, and under the shield of the _Pax Romana_, the towns were as open as now. And as little as now did they look forward to a time when each would have to become a strongly-held place of arms girded in by ma.s.sive ramparts, yet destined to prove all too weak against the sweep of barbarian invasion.

C. 8.--On most of these sites continuous occupation for many subsequent ages has blotted out the vestiges of their Roman day. Every town has a tendency literally to bury its past; and the larger the town the deeper the burial. Thus at London the Roman pavements, etc.

found are some twenty feet below the present surface, at Lincoln some six or seven, and so forth. To learn how a Roman town was actually laid out we must have recourse to those places which for some reason have not been resettled since their destruction at the Anglo-Saxon conquest, such as Wroxeter and Silchester, where the remains accordingly lie only a foot or two below the ground. The former has been little explored, but the latter has for the last ten years been systematically excavated under the auspices of the Society of Antiquaries, the portions unearthed being reburied year by year, after careful examination and record.[224]

C. 9.--The greater part of the site has thus been already (1903) dealt with; proving the town to have been laid out on a regular plan, with straight streets dividing it, like an American city, into rectangular blocks. Twenty-eight of these have, so far, been excavated. They are from 100 to 150 yards in length and breadth, arranged, like the blocks in a modern town, with houses all round, and a central s.p.a.ce for gardens, back-yards, etc. The remains found (including coins from Caligula to Arcadius) prove that the site was occupied during the whole of the Roman period. Originally it was, in all probability, one of the towns built for the Britons by Agricola[225] on the distinctive Roman pattern, with a central forum, town hall, baths, temples, and an amphitheatre outside the city limits.

C. 10.--The forum was flanked by a vast basilica, no less than 325 feet in length by 125 in breadth, with apses of 39 feet radius.[226] A smaller edifice of basilican type is generally supposed to have been a Christian church. It stands east and west, and consists of a nave 30 feet long by 10 broad, flanked by 5-feet aisles, with a narthex of 7 feet (extending right across the building) at the east end, and at the west an apse of 10 feet radius, having in the centre a tesselated pavement 6 feet square, presumably for the Altar.[227]

C. 11.--The main street of Silchester ran east and west, and _may_ have been the main road from London to Bath; while that which crosses it at the forum was perhaps an extension of the Icknield Way from Wallingford to Winchester. A third road led straight to Old Sarum,[228] and there may have been others. Silchester lies about half-way between Reading and Basingstoke.

C. 12.--The relics of domestic life found indicate a high order of peaceful civilization. Abundance of domestic pottery (some of it the glazed ware manufactured at Caistor on the Nen), many bones of domestic animals (amongst them the cat),[229] finger-rings with engraved gems, and the like, have been discovered in the old wells[230] and ashpits. More remarkable was the unearthing (in 1899) of the plant of a silver refinery,[231] showing that the method employed was a.n.a.logous to that in vogue amongst the j.a.panese to-day, and that bone-ash was used in the construction of the hearths.[232]

The houses were mainly built of red clay (on a foundation wall of flint and mortar) filled into a timber frame-work and supported by lath or wattle. The exterior was stamped with ornamental patterns, as in modern "parjetting" (which may thus very possibly be an actual survival from Roman days). This clay has in most cases soaked away into a mere layer of red mud overlying the pavements; but in 1901 there was unearthed a house in which a fortunate fire had calcined it into permanent brick, still retaining the parjetting and the impress of wattle and timber. But the whole site has not provided a single weapon of any sort or kind, and the construction of the defences clearly shows that they formed no part of the original plan on which the place was laid out.[233] They were probably, as we have said, added at the break up of the Pax Romana.

C. 13.--With the exception of the silver refinery above mentioned, nothing has appeared to tell us what handicrafts were practised at Silchester; but such industries formed a noteworthy feature of Romano-British life. Naturally the largest traces have been left in connection with that most imperishable of all commodities, pottery.

The kilns where it was made are frequently met with in excavations; and individual vases, jugs,[234] cups, and amphorae (often of very large dimensions) constantly appear. Many of these are beautifully modelled and finished, and not unseldom glazed in various ways. But there is no evidence that the delicate "Samian" ware[235] was ever manufactured in Britain, though every house of any pretensions possessed a certain store of it. The indigenous art of basket-making[236] also continued as a speciality of Britain under the Romans, and the indigenous mining for tin, lead, iron, and copper was developed by them on the largest scale. In every district where these metals are found, in Cornwall, in Somerset, in Wales, in Derbyshire, and in Suss.e.x, traces of Roman work are apparent, dating from the very beginning of the occupation to the very end. The earliest known Roman inscription found in Britain is one of A.D. 49 (the year before Ostorius subdued the Iceni) on a pig of lead from the Mendips,[237]

and similar pigs bearing the Labarum, _i.e._ not earlier than Constantine presumably, have been dredged up in the Thames below London.[238] Inscriptions also survive to tell us of a few amongst the many other trades which must have figured in Romano-British life,--goldsmiths, silversmiths, iron-workers, stone-cutters, sculptors, architects, eye-doctors, are all thus commemorated.[239]

C. 14.--But then, as always, the life of Britain was mainly rural. The evidence for this unearthed in the Cam valley has already been spoken of, and in every part of England the "villas" of the great Roman landowners are constantly found. Hundreds have already been discovered, and year by year the list is added to. One of the most recent of the finds is that at Greenwich in 1901, and the best known, perhaps, that at Brading in the Isle of Wight. Here, as elsewhere, the tesselated pavements, the elaborate arrangements for warming (by hypocausts conveying hot air to every room), the careful laying out of the apartments, all testify to the luxury in which these old landlords lived. For the "villa" was the Squire's Hall of the period, and was provided, like the great country houses of to-day, with all the best that contemporary life could give.[240] And, like these also, it was the centre of a large circle of humbler dependencies wherein resided the peasantry of the estate and the domestics of the mansion.[241] The existence amongst these of huntsmen (as inscriptions tell) reminds us that not only was the chase, then as now, popular amongst the squirearchy, but that there was a far larger scope for its exercise.

Great forests still covered a notable proportion of the soil (the largest being that which spread over the whole Weald of Suss.e.x)[242], and were tenanted by numberless deer and wild swine, along with the wolves, and, perhaps, bears,[243] that fed upon them.

C. 15.--Hence it came about that during the Roman occupation the British products we find most spoken of by cla.s.sical authors are the famous breeds of hunting-dogs produced by our island. Oppian[244]

[A.D. 140] gives a long description of one sort, which he describes as small [Greek: _baion_], awkward [Greek: _guron_], long-bodied, rough-haired, not much to look at, but excellent at scenting out their game and tackling it when found--like our present otter-hounds. The native name for this strain was Aga.s.seus. Nemesia.n.u.s[245] [A.D. 280]

sings the swiftness of British hounds; and Claudian[246] refers to a more, formidable kind, used for larger game, equal indeed to pulling down a bull. He is commonly supposed to mean some species of mastiff; but, according to Mr. Elton[247] mastiffs are a comparatively recent importation from Central Asia, so that a boarhound of some sort is more probably intended, such as may be seen depicted (along with its smaller companion) on the fine tesselated pavement preserved in the Corinium Museum at Cirencester.[248] Whatever the creature was, it is probably the same as the Scotch "fighting dog," which figures in the 4th century polemics as a huge ma.s.sive brute of savage temper[249]

and evil odour,[250] to which accordingly controversialists rejoice in likening their ecclesiastical opponents.[251] Jerome incidentally tells us that "Alpine" dogs were of this Scotch breed, which thus may possibly be the original strain now developed into the St. Bernard.

C. 16.--But the existence of such tracts of forest, even when very extensive, is quite compatible (as the present state of France shows us) with a highly developed civilization, and a population thick upon the ground. And that a very large area of our soil came to be under the plough at least before the Roman occupation ended is proved by the fact that eight hundred wheat-ships were dispatched from this island by Julian the Apostate for the support of his garrisons in Gaul. The terms in which this transaction is recorded suggest that wheat was habitually exported (on a smaller scale, doubtless) from Britain to the Continent. At all events enough was produced for home consumption, and under the shadow of the Pax Romana the wild and warlike Briton became a quiet cultivator of the ground, a peaceful and not discontented dependent of the all-conquering Power which ruled the whole civilized world.

C. 17.--In the country the husbandman ploughed and sowed and reaped and garnered,[252] sometimes as a freeholder, oftener as a tenant; the miller was found upon every stream; the fisher baited his hook and cast his net in fen and mere; the Squire hunted and feasted amid his retainers (who were usually slaves); his wife and daughters occupied themselves in the management of the house. The language of Rome was everywhere spoken, the literature of Rome was read amongst the educated cla.s.ses; while amongst the peasantry the old Celtic tongue, and with it, we may be sure, the old Celtic legends and songs, held its own. Intercourse was easy between the various districts; for along every great road a series of posting-stations, each with its stud of relays, was available for the service of travellers. In the towns were to be found schools, theatres, and courts of justice, with shops of every sort and kind, while travelling pedlars supplied the needs of the rural districts. No one, except actual soldiers, dreamt of bearing arms, or indeed was allowed to do so,[253] and the general aspect of the land was as wholly peaceful as now. But every one had to pay a substantial proportion of his income in taxes, in the collection of which there was not seldom a notable amount of corruption, as amongst the publicans of Judaea. In the bad days of the decadence this became almost intolerable;[254] but so long as the central administration retained its integrity the amount exacted was no more than left to every cla.s.s a fair margin for the needs, and even the enjoyments, of life.

SECTION D.

The unconquered North--Hadrian's Wall--Upper and Lower Britain--Romano-British coinage--Wall of Antoninus--Britain Pro-consular.

D. 1.--The weak point of all this peaceful development was that the northern regions of the island remained unsubdued. It was all very well for the Roman Treasury, with true departmental shortsightedness, to declare (as Appian[255] reports) that North Britain was a worthless district, which could never be profitable [Greek: [_euphoron_]] to hold. The cost would have been cheap in the end. All through the Roman occupation it was from the north that trouble was liable to arise, and ultimately it was the ferocious independence of the Highland clans that brought Roman Britain to its doom. The Saxons, as tradition tells us, would never have been invited into the land but for the ravages of these Picts; and, in sober history, it may well be doubted whether they could ever have effected a permanent settlement here had not the Britons, in defending our sh.o.r.es, been constantly exposed to Pictish attacks from the rear.