/^

BELG£.

FROM THE MONUMENT OF JOVINUS AT REIMS.

'

FRANKS.

FROM TERRA COTTA FIGURES FOUND NEAR NVMEGEN AND XANTEN.

After Lubach

Vrffc:

THE

A CONTRIBUTION TO

Gbc Hnthropoloo^ of Western Europe

JOHN BEDDOE

M.D., F.R.S., &c.

Ex-President of the Anthropological Society of London and of the Bristol Naturalists' Society

Foreign Associate of the Anthropological Society of Paris

Corresponding Member of the Anthropological Society of Berlin

Honorary Member of the Anthropological Societies of Brussels and of Washington

and of the Philosophical Institute of Bath

\% (pHHH

izl

BRISTOL J. W. ARROWSMITH, QUAY STREET

LONDON TRUBNER AND CO., LUDGATE HILL

I8»5

All rights reserved

TO

1R u t> o I f ID 1 r c b o w

AND

Paul ftopinarfc

AND

TO THE MEMORY OF

paui Broca

AND

3oscpb Barnarfc "©avis

This Work is Dedicated.

PREFACE

THE present volume is to a great extent an expansion or development of a manuscript Essay, which in 1868 carried off the great prize of the Welsh National Eisteddfod. This prize, consisting of one hundred guineas, contributed by an accomplished and public -spirited Welshman, Mr. A. Johnes, of Garthmyl, and supplemented by a promise of fifty more from the funds of the Eisteddfod Committee, had been competed for, without success, during four successive years, by numerous candidates, two at least of whom — Mr. L. Owen Pike and Dr. Nicholas — had published their essays.

The successful work, however, though composed expressly for the occasion, was really the outcome of a great part of the leisure of fifteen years devoted to the application of the numerical and inductive method to the ethnology of Britain and of Western Europe. Their fruit, though satisfactory to the accomplished arbiter (the late Lord Strangford), did not appear to me ripe for publication. Having since then added largely to my material, and accumulated as great a store of observations as my opportunities are likely to afford, I now offer my conclusions and opinions to my brother anthropologists and to the public. In doing so, it has seemed best to present to them also the greater part of the facts and figures on which these opinions have been founded. Those relating purely to stature and bulk are not, however, republished ; as they have been the subject of a monograph in the Anthropological

VI PREFACE.

Memoirs, and have also been extensively utilized by Mr. Charles Roberts and Sir Rawson Rawson, in the Anthropometric Reports of the British Association.

Since the publication of my Contribution to Scottish Ethnology, one of the earliest pieces of work in this field, in 1853, the scientific value of numerical observations on colour as well as stature, conducted on a large scale, has come to be generally acknowledged ; and under Virchow and Ranke in Germany, Vanderkindere in Belgium, Meyer and Kopernicki in Poland, Kollmann, Guillaume, and others in Switzerland, and Guibert in Brittany, statistics of great extent and value, and arranged more or less on the same plan as my own, have been accumulated and published ; while in France, Topinard hopes to induce the Govern- ment to carry out, on a national scale, and on my system, a similar investigation.

The ever-increasing rapidity of local migration and intermixture, due to the extension of railways and the altered conditions of society, will in the next generation almost inextricably confuse the limits and proportions of the British races ; and it is a source of satisfaction to me that I have laboured to seize on fleeting opportunities, and to observe and record phenomena, which, how- ever trivial they may appear from some points of view, may for generations to come retain some biological and historical value.

I owe a heavy debt of gratitude to many friends, and to some whom I do not even personally know, for assistance in the work ; but the only one I can here mention is Mr. David Davies, who has lent his valuable aid in the revision of the proofs, during my absence from England. Most of the drawings were executed by Mr. Norman Hardy.

Hablc of Contents.

Chap. Page.

I. On Methods ......... i

Uncertainty of casual observation and current opinion as to physical charac- teristics— Importance of colour in race-type — Difficulties connected with personal equation — Arrangement of colours finally adopted — Comparison thereof with that of Broca — Method of working — Index ot Nigrescence — Deficiency of observations on modern and mediaeval British skulls — Lamentable destruction of ossuaries — Measurement of the living head — Method of noting facial characters.

II. Prehistoric Races ........ 9

Palaeolithic Races — Probable vestiges in modern population — Mongoloid type — Afiicanoid type — The Neolithic period — Long-barrow skulls — Perth-y-chwaren type: its mesocephaly — Northern and Southern long-heads — Influence of "media" on colour — Round-barrow or Bronze race — Its probable connexion with the Borreby and Sion types or races, and with the Walloons — Its presence in our modern population — Pear-shaped skull, the prevailing one among the pre-Roman population.

III. Britain before Caesar and Claudius, etc. ... 19

Languages of the Britons — Ethnological position of the Belgae — The Monument of Jovinus — Vanderkindere's statistics, and my own — Treves and the Galatians — The Coranied and Coritavi — The Caledonians — The Firbolg — The Lloegrians and Iberians — Gaels in Great Britain — The Brythons — Summary.

IV. The Roman Period ........ 30

Ethnological changes during this period — Magnitude and character of the Roman immigration — Subsequent revival of British nationality — The Roman Wall — The Gaels in Wales — Movements in Ireland — The Litus Saxonicum — Departure of the Romans.

V. The Anglo-Saxon Conquest and Period ... 38

The Saxons, Angles, Jutes, and Frisians — Their local origin and relations — Frisian head-form — Gildemeister's researches at Bremen — The Old Saxon forms — Theory of the English or Anglo-Saxon Conquest — Estimate of authorities — Formation of the new kingdoms, and successive extensions at the expense of the Britons — The Picts, Caledonians, and Scots — Ethnological character of the populations sub- sequently to these invasions — Evidence of proper names — The word Wealh — The Will of King Alfred — Legal systems — Roman, Celtic, or Teutonic Cavelkind — The Churls — The constitution of the Manor — The fate of the Romano-British towns — The Bridge and Burgh Rate — Extinction of Christianity — The Sixhyndman — How far did "the Lloegrians become Saxons"? — Theory of the early history of Mercia — Evidence of local names in England and Scotland — Evidence of current language — Vowel sounds — British derivatives — Spindle- and servile-words — The Rhyming. score — Reasoning from language of less import as one travels westward.

VI. Germanic Conquests elsewhere, especially in Switzer- land .......... 72

Modern ethnology of Flanders, Brabant, and Treves — Points of agreement and of difference between the Saxon conquest of England, and the Alemannic of Switzerland — Brief notice of the Burgundians — Alemannic, Swabian, and Rhastian types — Anthropological statistics of school-children in Switzerland — Kollmann and Studer thereupon — Guillaume on the darkening of hair with age, and Mayr on town populations — Dunant on stature of Swiss — Questions of method and mapping — Map of Index of Nigrescence &c, from official statistics — Map of the same from personal observation — Nidwalden Grisons — Decline of the blond long-headed type — Ranke's observations on the Bavarians — Summary and conclusions — Table of personal observations.

VII. The Danish Period ....... 86

Classification of the Scandinavian settlements and operations — Numbers of the invaders — The Danelagh — Western settlements — Ethnological history of Southern Cambria and the Isle of Man — Infiltration elsewhere — Saxonization of the West of England — The "Dunsetas" — The Welsh Marches — The Scandinavian elements and history in Scotland and Ireland.

Vlll TABLE OF CONTENTS.

Chap. Page.

VIII. The Normans 93

Racial character of the Norman immigration and of the populations which supplied it — Difference between the ruling and subordinate classes in Normandy — The Bretons, " French," and Flemings.

• IX. The Norman Conquest ....... 97

Main facts of its history, and of the so-called "Saxon Conquest" in Scotland, and the Anglo-Norman one in Wales — Effect of the Conquest on p distribution of race — Evidence of Domesday imperfections therein I i numbers of the several castes: their status — Position of the native English in Kent, as deduced from Dugdale — The same in Essex, Suffolk, Norfolk, Lincolnshire, Hampshire, Gloucestershire, &c. — Evidence from names of witnesses to charters — Status of the race-elements in Northumbria, exclusive of Yorkshire.

X. The Normans in Yorkshire . . . . . .110

Destruction of the Anglo-Danish population by sword, pestilence, and famine — Local exceptions — Criteria of the actual loss — Map and details of Richmondshire — Return home of the Bretons — French colony in York — Survival of some native Nobles as tenants, unnoticed in Domesday — Meaux Abbey — General conclusions — Analytical Table of population mentioned in Domesday Book for Yorkshire and other counties.

XI. Norman-French Immigration . . . . . .119

Immigration subsequent to the Conquest— The Norman People— Systems of Personal Nomenclature — Interma riage — Settlement of Englishmen in Normandy — List of house-owners in Winchester — List of names in the B and in the

Liber Vita of Durham— Evidence of the Hundred Rolls— Analytical Tables of Personal Names in the 12th and 13th centuries— Analytical Table of Modern English Surnames — Examination of the Hundred Rolls: their great value — Camden on changes of surnames — Errors of the author of The Norman People — Examination of my own methods — Inferences as to proportion of " Norman " blood in England.

XII. Subsequent Migrations 136

Decline of the Scandinavians in Ireland — Anglo-Norman and English invasions and colonizations therein — Evidence of surnames as to proportion of "Celtic" and "Saxon" blood — Remains of the Norsemen in the Hebrides, &c. — The modern Irish immigration into Scotland — The Huguenots and Palatines — Immigration of Scotch- men, Irishmen, and Welshmen into England : its amount and distribution.

XIII. Preface to the Tables and Maps, Considerations on

Methods of Computation, and of Division of Types 143

Military Statistics, and Schedules of Eye and Hair-colour — Comparison and contrast thereof with those of the Anthropometric Committee — General corres- pondence of the former with those of the Author — Characteristics of the descendants of Irish immigrants into Great Britain — Brief notices of other Tables.

XIV. General Commentary on the Tables .... 239

Natives of the different Scotch islands, description and origin of — Highlanders, difficulty of classifying, owing to their being a heterogeneous race ; description of their central type; the dolicephalous and brachycephalous Celt — The' red-haired Gael and the Iberian — The Angles and Anglo-Danes in Scotland — Mental and moral attributes of the Southern Scots — Northumberland and Berwick strongly Anglian — Durham, old race Anglian and Danish — Yorkshire: North and East Ridings, Anglo- Danish — A clear and distinct moral type found in Yorkshire: description of — Ethnology of North Lancashire, Northumberland and Cumberland — Nottingham- shire and Lincolnshire Anglo-Danish Counties: description of Danish type by Macintosh — Derbyshire, physical type Anglian— East Staffordshire Anglian, but not Danish — Leicestershire colonized by Danes, but retains many of the dark pre- Anglian stock — Ethnology of Midland and South-midland Counties of Essex, Middlesex, Kent and Surrey — Isle of Wight, type Jutic — New Forest, primitive population remaining — Devonshire, ethnology of; singular beauty of the women — Cornwall, description of natives — West of England, including Bristol, population of — Wales, description of physical and moral characteristics of natives ; not a homogeneous race — Ireland has one distinct type, easy of description — Physiognomy dependent on form of skull — Description of local sub-types and their probable origin — A numerical expression of the complexions or colour of the skin in several parts of the British Islands.

Final. Conclusions and Inconclusions ..... 269

Value of the index of nigrescence — General estimate of the result of great ethnological movements— The tendency of the darker races to swamp the blond Teutons of England by a reflux migration — A short enumeration of ethnological points of great interest which still remain unsettled — Aim in writing, and conclusion.

THE RACES OF BRITAIN.

CHAPTER I .

©n flDetbobs.

IT was the ancient controversy respecting the colour of the hair of the Kelts, then burning briskly enough, and even now still smouldering, that led me to begin systematic numerical observa- tions in physical anthropology. Very little reading sufficed to show me that, if it was a difficult task to ascertain the complexional peculiarities of the Kelts of 2000 years ago, it was a no less puzzling one to deter- mine those of their supposed representatives at the present day. It is of little use to appeal to current opinion, or the results of casual obser- vation. The eye may rest upon a great many sets of features in the course of a long day's travel, but the mind will retain but few of them photographed on the tablet of memory ; and those few will probably be such as have presented striking peculiarities, or have belonged to the persons brought most frequently and nearly in+o the company and con- tact of the observer. This fact, together with the inveterate tendency of so many scientific observers, to see everything as they wish and expect it to be, rather than as it is, may account for the striking discrepancies among ethnological writers on this simple matter of fact. Thus it comes to pass that some attribute blue, others black eyes to the majority of the Irish ; some describe the men of Kent as particularly fair, others as " assimilated to the dark-complexioned inhabitants of the opposite coast." The minister of Wick, in his Statistical Account of the place, described his parishioners as "having for the most part dark brown or black hair, and dark complexions ; remarkably few having red or yellow hair." My own impressions on visiting the town were quite of an opposite nature, and were confirmed by an enumeration of the com- plexional characters of more than 300 individuals.

Similar discrepancies are manifested whenever any one attempts to define the prevailing complexion of the continental Teutonic peoples, or of the Slavonic, or of almost any other race or tribe. Take, for example, the Croats. One of two observant travellers talks of their "shaggy

2

2 THE RACES OF BRITAIN.

black locks," and another of "their Slavonian characteristics of blue eyes and fair hair." I could quote two or three other descriptions of their persons, which only agree in differing every one from the other.

Taking note of all these contradictions, and considering, moreover, that the material to be worked upon lay ready to hand in streets and market-places, not hidden away in museums and charnel-houses, I resolved to make observations on colour phenomena on a large scale, so as to afford to anthropologists some trustworthy material whereon to build, or at least to test the permanence of characteristics which were by some treated as fugitive, by others as well-nigh everlasting. How far I have succeeded in the first-mentioned object it will be for such of my readers to determine who shall have borne with me to the end. With regard to the second, it is, doubtless, natural that a subject to which one has given much time and labour should gradually increase in apparent importance under one's eyes, and that he who has given more attention to colour than to form should be disposed unduly to exalt the value of the first. Be that as it may, I have come to estimate very highly the permanence of the colours of hair and eyes. It is, of course, impossible for an evolutionist to regard them as absolutely permanent. But one may readily conceive, as I do, that whenever a distinct and tolerably homogeneous breed has been established, its colours may remain very much the same so long as the conditions of natural selection remain nearly identical. There are naturalists of eminence who regard these characters as fugitive, nay, almost accidental, so to speak, compared with the form of bones, especially of the skull -bones. "Colour," I once heard Sir Henry Rawlinson say, at an ethnological gathering, " is no part of type." From this view I strongly dissent. In the same family, in the same tribe, variations in head-form are usually about as notable as variations in complexion. There is as much to show for Schaafhausen's notion, that civilisation tends to widen the head, as for another very prevalent idea, that it tends to darken the hair. Indeed, it is more easy to reckon up agencies which might operate on the head- form than such as might alter the colour of the iris. Of the latter, the only feasible one that occurs to me is an increase or diminution in the amount of light to which the organ is habitually exposed ; * but the former are many, e.g., changes in the soil, or in the food produced there- from, or in the character of the diet, increasing or lessening the supply of phosphate of lime; t changes in civilisation, involving greater or less employment of certain portions of the brain ; and so forth.

Unfortunately we have not the same kind of evidence respecting colour in past ages as we have respecting form. We know by the

* Thus natural selection, in a dark, cloudy climate like that of the West of Ireland, may have tended to lighten the colour of the eye, the protection of much dark pigment being unnecessary.

t See Durand de Gros on the population of the Aveyron, in the Bulletins de la Societe d' Anthropologic

THE RACES OF BRITAIN. 3

evidence of our own eyes that flattened leg-bones and perforated humeri were common in certain ancient races, and that they are now very uncommon in their supposed descendants, or at any rate in the present occupiers of the same countries. But as to whether red hair was more common then than now, we cannot have the same assurance : such hair as has come down to us from individuals of ancient races is generally stained and altered, so as to be untrustworthy evidence ; and the representations of colour on walls and vases are evidently conventional, and afford, at the best, only material for inferential argument. Finally, the descriptive statements of ancient geographers and historians, even when obviously intended to be careful and accurate, are liable to two qualifications, one of which is our ignorance of the precise shade of meaning their contemporaries and countrymen attached to certain adjectives of colour, and the other is our ignorance of the personal equation of the observer, the latter objection, of course, applying equally to the statements of modern travellers and naturalists, whose nomenclature of colours and shades often differs very widely. Thus almost all French anthropologists say that the majority of persons in the north of France are blond ; whereas almost all Englishmen would say they were dark, each set of observers setting up as a standard what they are accustomed to see around them when at home. What is darkish brown to most Englishmen would be chestnut in the nomen- clature of most Parisians, and perhaps even blond in that of Auvergne or Provence ; an ancient Roman might probably enough have called it suffiavus, or even flavus. Nor is this difference of personal equation confined to observers who differ in nationality, or who dwell among populations which differ strikingly from each other in colour. The plan for obviating this difficulty devised by the illustrious Broca, though very useful as regards the hues of skin and irides, is less so for those of the hair, which are not flat tints ; and the scale of hair-colours contrived for the Anthropometric Committee of the British Association by General Pitt-Rivers, was found of little practical use for the same reason.

My first observations were vitiated by faulty classification ; but I soon settled down into the system to which I have since adhered, and which recommended itself chiefly by its convenience, as it generally enabled me to locate an individual in his proper class and division on a very cursory inspection.

I acknowledge three classes of eyes, distinguished as much by shade as by colour — light, intermediate or neutral, and dark. To the first class are assigned all blue, bluish -gray and light gray eyes (caerulei, cinereo-coerulescentes, coesii). These correspond to the five blue types of Broca, n, 12, 13, 14, 15, and to No. 10, his lightest green. It was perhaps a mistake to include the darkest blue (No. 11) in this category, as in an unfavourable light it is liable to be confounded with " black."

In the third class I put the so-called black eyes, and those usually called brown and dark hazel. These correspond to the deeper

9 *

4 THE RACES OF BRITAIN.

shades of Broca's orange, green, and violet-gray, Nos. I, 2, 3, 6, 7, 16, 1.7, 18.

To the second, or neutral class, remain dark gray, brownish-gray, very light hazel or yellow, hazel-gray, formed by streaks of orange radiating into a bluish -gray field, and most shades of green, together with all the eyes of whose colour I remain uncertain after an ordinarily close inspec- tion. These correspond pretty exactly, I believe, to 4, 5, 8, 9, 19, and 20.

Each of my three classes of eyes is sub-divided into five, in accord- ance with the accompanying colour of hair :

Class R includes all shades which approach more nearly to red than to brown, yellow, or flaxen.

Class F (fair) includes flaxen, yellow, golden, some of the lightest shades of our brown, and some pale auburns in which the red hue is not very conspicuous.

Class B includes numerous shades of brown, answering nearly, I believe, to the French chatain and chatain-clair, but perhaps less extensive on the dark side.

Class D corresponds nearly with the French brun, most of their brun-fonces, and the darkest chatains, and includes the remaining shades of our brown up to

Class N (niger), which includes not only the jet-black, which has retained the same colour from childhood, and is generally very coarse and hard, but also that very intense brown which occurs in people who in childhood have had dark brown (or in some cases deep red) hair, but which in the adult cannot be distinguished from coal-black, except in a very good light.

When unable to decide in which of two columns {e.g., B or D) an individual ought to be inscribed, I divide him between the two, by a Solomonian judgment, and set down \, or -5, in each of them.

When engaged in this work I set down in his proper place on my card of observation every person (with the exceptions to be mentioned presently) whom I meet, or who passes me within a short distance, say from one to three yards. As a rule, I take no note of persons who apparently belong to the upper classes, as these are more migratory and more often mixed in blood. I neglect those whom I suppose to be under age — fixing the point roughly at 18 or 20 for men, 17 or iS for women — as well as all those whose hair has begun to grizzle. Thus I get a fairly uniform material to work upon, though doubtless the hair of most people does darken considerably between 20 and 40 or 50. In order to preserve perfect fairness, I always examine first, out of any group of persons, the one who is nearest, rather than the one to whom my attention is most drawn. Certain colours of the hair, such as red, certain shades of the eye, such as light gray, can be discerned at a very considerable distance ; but I take no note of anyone who does not approach me so nearly that I can recognise the more obscure colours. Much allowance needs to be made for the varying effects of light.

THE RACES OF BRITAIN. 5

Direct sunlight is better avoided when possible ; I always choose the shady side of a street on a sunny day. Considerable difficulties are created by the freaks of fashion. I once visited Friesland, in order to study the physical type of that region. Conceive my disappointment when I found myself surrounded by comely damsels and buxom matrons, not one of whom suffered a single yellow hair to stray beyond her lace cap or silver-gilt head-plate. When I began to work in England dark hair was in fashion among the women ; and light and reddish hues were dulled with greasy unguents. In later years fair hair has been more in vogue ; and golden shades, sometimes unknown to nature, are produced by art. Among men, on the other hand, the close cropping of the head, borrowed from the French, makes comparisons difficult. Fortunately, most vagaries of this kind are little prevalent in the classes among whom I seek my material.

It may be objected that there is no security that many of the persons observed may not be aliens to the place or neighbourhood wherein they are encountered. Certainly, there is no such security. But if a sufficient number of observations be secured, and the upper and other notoriously migratory classes (who are mostly easy of recognition) be excluded, the probability is immense that the great majority of the remainder have been born within a moderate radius of the centre of observation ; and the majority will determine the position of the community in my chro- matic scale.

A ready means of comparing the colours of two peoples or localities is found in the Index of Nigrescence. The gross index is gotten by subtracting the number of red and fair-haired persons from that of the dark-haired, together with twice the black-haired. I double the black, in order to give its proper value to the greater tendency to melanosity shown thereby ; while brown (chestnut) hair is regarded as neutral, though in truth most of the persons placed in B are fair-skinned, and approach more nearly in aspect to the xanthous than to the melanous variety.

D + 2N-R-F = Index.

From the gross index the net, or percentage index, is of course readily obtained.

It must not be supposed that, in devoting so much time and care to the collection of facts relating to colour, I was influenced by any exces- sive estimate of their importance. My chief inducement was the great abundance of the material, which, from a scientific point of view, was running to waste. The same thing might indeed be said of the heads of the British population ; for they were also generally neglected by ethnologists, whatever phrenologists might be doing, the former being almost entirely absorbed in ancient craniology. But there was a very important difference between these two lines of enquiry : the one could be pursued without the concurrence of the subjects, the other could not.

6 THE RACES OF BRITAIN.

Had there been anything like a complete craniological record, had there even been anything approaching the amount of ancient and mediaeval material that can be used in France or Switzerland, one might have neglected the heads of one's contemporaries, in consequence of the obstacle just mentioned ; but in truth the record is anything but com- plete or satisfactory, notwithstanding the exertions of Davis and Thurnam, of Bateman, of Greenwell and Rolleston, of Daniel Wilson, James Hunt, and Pitt -Rivers. This lamentable defect arises partly from the destructive ignorance of our earlier antiquaries, who, while they carefully collected every fragment of a potsherd from the barrows they explored, utterly neglected, and exposed to decay, the often more important osseous remains. Even now " finds " frequently occur, the benefit of which is lost to anthropology, from the absence of qualified observers, and the lack of knowledge or interest in the finders and their neighbours. It is not long since there existed several mediaeval ossuaries in England, systematic observations in which might have been of some value ; but with the exception of those at Hythe, Rothwell, and Michel- dean, they have all, I fear, been destroyed. Thus a very fine one at Ripon was destroyed, unmeasured and undescribed, by the late Dean Macneile, and another in the crypt of Tamworth was turned out to make room for a heating apparatus. These misfortunes are the more to be regretted, inasmuch as we really do not possess sufficient osseous material in our museums for determining the form and size of the skull of the modern Englishman. The few we have are in great part those of criminals, lunatics, and paupers. In this respect, owing to a prejudice, from some points of view respectable, we are behind most European nations ; and when, in such works as those of Topinard and De Quatrefages, we see comparisons drawn between the ordinary skull- forms of different countries, England is usually conspicuous by its absence.

On account of this dearth of material, I have measured a consider- able number of living British heads, and shall make use of the results of these measurements in the present volume. As no accredited method existed when the work was begun, it was necessary to frame one. The difficulties in the way were considerable, and certainly were but partially overcome. It was necessary to avoid fatiguing or irritating the subjects ; yet it was desirable to obtain as many data as possible suitable for com- parison with those taken from ancient crania. With much regret I abandoned the use of Mr. Busk's excellent craniometer, and with it all radial measurements, because it sinned against the former of these re- quirements, and restricted myself to the use of the index callipers and graduated tape.

There are few points on the living head that are positively identifi- able ; and I was compelled to retain the use of some which are open to the objection of not being so. Some of the tape measures are affected by the variations in quantity and length of hair, though to a less extent

THE RACES OF BRITAIN. 7

than might be supposed. The following are those which I have been accustomed to take :

A . With the callipers, (a) Lengths.

i. Maximum length from the glabella.

2. Length from the inion or occipital tuberosity to the most

prominent part of the frontal curve.

3. Glabello-inial length.

4. Maximum length according to Barnard Davis, i.e., from

the ophryon, or the flat space above the glabella.

By the use of these four it is possible, in pursuance of one of Broca's suggestions, to work out the degree of prominence of the occiput, of the forehead, and of the glabella, and thus to compensate in some degree the lack of radial measurements.

5. Vertico-mental length, or maximum length of the whole

head from crown to chin.

(b) Breadths.

6. Frontal minimum breadth, just above the brows.

7. Breadth at the stephanion, or maximum frontal. This is

very uncertain ; in many heads it is impossible to be sure whereabout the stephanion is.

8. Zygomatic breadth, maximum.

9. Auricular breadth, gotten, in accordance with Broca's

recommendation, at the pit just in front of the helix, and above the condyle of the jaw and the root of the zygoma. This is a valuable measurement, the point being so easily identifiable. In conjunction with No. 11, it yields information as to the breadth of the base of the skull.

10. Maximum breadth, wherever found, and where found.

11. Mastoid breadth. Taken at the most prominent part of

the external mastoid curve. This is very faulty, from the difficulty of fixing on the same point in different heads, the shape of the mastoid protuberances varying much. B. With the tape, (e) Circumferences.

12. Circumference in the line of length 1.

13. do. do. do. 2.

14. do. do. do. 3.

15. do. do. do. 4.

13, 14, and 15 are of comparatively little value, seldom adding any- thing to the information given by 1, 2, 3, 4, 9, and 10. I often omit them.

8 THE RACES OF BRITAIN.

\rcs.

16. From the nasal notch to the inion, or occipital tuberosity.

17. From opposite the centre of one auricular meatus to that

of the other, in a vertical line. This is very useful. Taken in connexion with 1 and 16, it gives a fair idea of the height of the head ; but it has disadvantages, of which the chief is the uncertainty of the vertical line.

18. From the centre of one meatus to that of the other, along

the superciliary ridges and the glabella. Indicates the comparative frontal and occipital development.

It is perhaps unfortunate, especially in view of the great attention now given to the facial bones, that I have seldom taken the length from the chin to the nasal notch, nor the breadth nor length of the nose. It has been my custom, however, to sketch the facial portrait by means of a few initial letters. Thus F., Sc, Pr., Br., Aq., Si., Ang. sketches a man of Fair complexion, Scutiform face, with Prominent brows, Broad cheekbones, Aquiline and Sinuous nose, and Angular chin.

I have spoken of the necessity and frequent difficulty of obtaining the consent of the owner of the head to be examined. His reluctance may sometimes be overcome by means of money, without going to the extent of the new hat always jocularly demanded in such cases. Some- times other means have proved successful. I cannot resist detailing those by which I succeeded in obtaining a valuable series of head- measurements in Kerry. Our travelling party consisted of Dr. Barnard Davis, Dr. T. Wise, Mr. Windele, and myself. Whenever a likely little squad of natives was encountered the two archaeologists got up a dis- pute about the relative size and shape of their own heads, which I was called in to settle with the callipers. The unsuspecting Irishmen usually entered keenly into the debate, and before the little drama had been finished were eagerly betting on the sizes of their own heads, and begging to have their wagers determined in the same manner.

CHAPTER II.

prehistoric IRaces.

THANKS to Boyd Dawkins, and Lubbock, and Evans, and a number of other workers at home and abroad, we know, or at least conjecture with some confidence, a great deal about the surroundings and mode of life of palaeolithic man in Western Europe — nay, even about his intellectual development, on which the very spirited and even artistic drawings and carvings he has left us throw some little light. But about his physical type we know next to nothing. Nothing like a strong case, so far as I am aware, has yet been made out for the palaeolithic position of any ancient skull discovered in England. The scepticism which some English anthropologists extend to the position of crania generally reputed palaeolithic by continental writers — e.g., those of Cro-Magnon on the one hand, and those of Furfooz on the other — may, perhaps, be overstrained ; but though England was, doubtless, a part of the continent of Europe at the date generally assigned to these skulls, it would not necessarily follow that the race even of Furfooz should have extended its habitat into this country.

If our palaeolithic race were really the ancestors of the Eskimos, or at least their near relations, as Boyd Dawkins would have them to be, it is at least possible that they may have left descendants behind them to mingle their blood with the neolithic races and their descendants of to-day. Now I think some reason can be shown for suspecting the existence of traces of some Mongoloid race in the modern population of Wales and the West of England.

Their most notable indication is the oblique or Chinese eye, with its external angle in a horizontal plane a little higher than the internal one. This is usually accompanied by an almond-like form of the opening, and a peculiar thickness of the upper eyelid : these latter characters may occur without the obliquity of the opening, but with a physiognomy referable to the same type.

I have notes of 34 persons with oblique eyes. Their heads include a wide range of relative breadth, from 72 to 86*6 ; and the average index of latitude is 78"9, which is not much greater than the average of England and Wales. But in other points the type stands out distinctly. The cheek-bones are almost always broad ; the brows oblique, in the same direction as the eyes ; the chin, as a rule, narrow

IO THE RACES OF BRITAIN.

or angular ; the nose is often concave or flat, seldom arched ; and the mouth is rather inclined to be prominent. The forehead usually recedes a little ; the inion is placed high, and the naso-inial arc is rather short (13*8 inches), so as to lead one to suppose that the cerebellum is scarcely covered by the posterior lobes. The iris is usually hazel or brown, and the hair straight, dark brown, black, or reddish. This type seems to be common in Wales, in West Somerset, and especially in Cornwall.

Colour of Eyes — Six blue or blue-gray, 1 gray, 3 dark gray, three hazel-gray or neutral brown, 16 hazel, 5 brown.

Colour of Hair — Two red, 1 red-brown, 1 dark chestnut, 6 brown, 4 darkish brown, 15 dark brown, 1 brown-black, 2 black.

Locality — Wales, 8 ; Dean Forest (Welsh surnames), 2 ; East Somerset, 2 ; Mid Somerset, 3 ; West Somerset, 6 ; Devon, 4 ; Cornwall, 5 ; other counties, 4.

No instance of this type has turned up among the (comparatively few) heads from the East of England which I have had opportunity for measuring, and very few from Ireland. I believe, however, that specimens of it might easily be found in the mountainous parts of Connaught, especially on the borders of Sligo and Roscommon. I have seldom noticed it in Scotland, but it occurs in Shetland. Dr. Mitchell mentions the obliquely-set eye in his description of one of his Scottish types, "the Irish Celt or Fin;" but though I am acquainted with the type he evidently had in mind, I cannot recognise in it an)- resemblance to the Finns of Finland, nor to the pattern of features just described.

There is an Irish type, known to Mr. Hector Maclean, and admirably described by him,* which I am disposed to derive from the race of Cro-Magnon, and that none the less because, like some other Irish types, it is evidently common in Spain, and furnishes, as Maclean remarks, the ideal portrait of Sancho Panza. It is said to be pretty common in the Hebrides, but rare in the Highlands. In the West of Ireland I have frequently seen it ; but it is curious, psychologically, that the most exquisite examples of it never would submit to measurement. Though the head is large, the intelligence is low, and there is a great deal of cunning and suspicion.

There are, however, in my lists more than 40 persons who are noted as prognathous, or, more exactly, "having prominent mouths;" 29 of these are English, 5 Welsh, and 11 Irish ; roughly speaking, about 6 per cent, of the English, 8 per cent of the Welsh, and 20 per cent, of the Irish list. The "Mongoloids" and the "prognathous" overlap each other in six instances ; but except in these cases there are very decided points of difference. The latter have longer and narrower heads ; their index of breadth is but 76-5, and in the bare skull would never exceed 80. The cheek-bones are much narrower (135 against 141 millimetres), but almost invariably prominent in the face. The usual form of the forehead is flat, narrow, and square; that of the chin, * Antkropol. Review, Vol. IV., p. 218.

THE RACES OF BRITAIN. II

narrow and often receding ; that of the nose, oftener concave than straight, oftener straight than sinuous or aquiline, usually prominent at the point, with the long slitty nostrils, which, whencesoever derived, are a characteristic of the modern Gaels. The flatness of the temporal region, which comes out in the narrowness of the diameter at the root of the zygoma, gives to the norma verticaolis that coffin- or pear-shape which Daniel Wilson ascribes to the Celts. :;c The hair is generally very dark and often curly, but the eyes are more often blue or light- or dark- gray than of any shade of brown ; they belong to the blue and violet scales of Broca rather than to the orange.

This is evidently the Gaelic type of Mr. D. Mackintosh,! whom I rank with Hector Maclean as one of the best observers and recorders of local physiognomy. Mackintosh finds these people very numerous in Dorset and Devon, especially towards Exmoor ; and several of my specimens came from that quarter. It may be worth notice that there was a large immigration from Ireland into North Devon in the sixteenth century, during one of the perpetually recurring seasons of civil strife in that island. I should think, however, that such immigrants were probably Anglo-Irish from the towns and coast-districts, and not Gaels at all, and that they returned to Ireland during the next interval of peace.

This one character of prognathism, taken separately, may be objected to as being of small value ; but there is, as I have shown, a very great similarity in other respects among the individuals who present it. It will scarcely do to ascribe it, as is often done, to the effect of misery and oppression on the physique of the race. The average stature of my 34 was 5ft. 7*6 in. (1717 metre): my material is taken mostly from the labouring classes, yet in the prognathous list appears one of the ablest and most distinguished clergymen in Wales. I have also noticed it in the portraits of some well-known Welsh bards ; in fact, eloquence, or at least readiness of speech, seems to be a general characteristic of the type.

While Ireland is apparently its present centre, most of its lineaments are such as lead us to think of Africa as its possible birthplace ; and it may be well, provisionally, to call it Africanoid, applying the name Atlantean, which has been suggested, to the widely- diffused Ibero- Berber race type, of which it is probably a subdivision, in spite of the wide difference in the form of the jaws between it and the Basque type of Zaraus, the best accredited Iberian standard.

Though I believe this Africanoid type to have been of very high antiquity, it must be acknowledged that we have no evidence carrying back its presence, in any of the British Isles, beyond the polished stone period. But the best authenticated ancient skulls from Ireland may have belonged to it ; for example, the three from the Phcenix Park tumuli (of which two are figured in the Crania Britannica), and those from the * Prehistoric Annals of Scotland. t Anthrop. Review, Vol. IV., p. 15.

12

THE RACES OF BRITAIN.

bed of the Nore at Borris." These show the inclination to prognathism to be of remote date in Ireland, as well as the peculiar form of low, straight brow that still prevails there, and which is connected with low, square, horizontal orbits.

COMPARATIVE TABLE OF MONGOLOID AND AFRICANOID TYPES.

	LIGHT BYES.		NEUTI	,AL EVES.
hair: a	Fair. Brown. Dark.	0 ri 5	0 h	Red. Fair.	Brown. Dark.	Black. Total.
Mongoloids (34)... 1	— 23	—	6	•5 —	2-5 2	— 5
Africanoids (35)... 1*5	2-5 6 8 3ARK EVES.		18		•5 2-5	1 4 U O g
hair: S 3 <& Ph	Brown. Dark.	"0	O H	Stature	Weight	c tic
Mongoloids... 1-5 —	4 x5	2'5	23	5 ft- 7 in	i5I-3	647
Africanoids... — —	i\5 7	4-5	13	5 ft. 7-6	in. i47'6	70
	HEADS, BREADTH. 1		Length.
Frontal 7 . Minimum. ^°	ic. Auricular.	Maximum.	Index.
Mongoloid (10) ... 4-25	5'67		5*54	6-o8	770	78-9
Africanoid (13) ... 4.20	5,29		5-i5	5'94	776	76-6

Gbe IReolitbic pcrto^♦

Since Daniel Wilson asserted the priority in Britain of the kymbe- kephalic or boat-shaped skull, and Thurnam broached his theory of " long barrows, long heads ; round barrows, round heads," so much evidence has been produced in favour of their views, and so little against them, that they may be regarded as fairly established. The frequent — nay, almost general — use of the tumuli and cairns of previous popula- tions by later comers raises a difficulty sometimes as to which may have been the primary, which the secondary interment ; and mistakes may have been made in many instances. We may probably, however, put almost entire trust in the list of skulls from long barrows and chambered tumuli, given in the Crania Britannica. This list was drawn up, I believe, by Dr. Barnard Davis, who, at the time of its construction, had not become a convert to the long-barrow hypothesis of his colleague, and, consequently, was free from any unconscious bias, which might have led him in some instances to weed out the broader heads as secondary interments. There are 31 skulls sufficiently perfect to yield the pro- portions of both breadth and height, as well as length ; and the * Figured in Laing and Huxley's Prehistoric Remains of Caithness.

THE RACES OF BRITAIN. 1 3

latitudinal index comes out as low as 71 -6, and the altitudinal at 72-6, the latter being taken, after the method of Ecker, from the plane of the occipital foramen, and being larger than the preferable method of Broca would make it.

Rolleston, operating on 10 skulls from the long barrows near Swell, in Gloucestershire, obtained average indices of 72*8 for breadth, and 76*5 for height ; and Garson, from 5 Oreadian skulls, which he attributes to the same period, gets a breadth index so low as 70*3. D. Wilson got from 9 of what he considered "primitive skulls" an index of 71*6; but his mode of selection was open to criticism.

But though these figures are probably adequate to give us a correct idea of the general form of cranium in the tenants of long barrows, it must not be taken for granted that they represent accurately the entire population of the neolithic period. We know well, from the light thrown on the ethnological history of Southern Germany by Ecker and Ranke and Von Holder, that it is quite possible for a large section of the population to be, during long periods, almost entirely unrepresented in the ordinary burial-grounds. The Mongoloid type, which I have already described, if it be really a race-type, and not merely a har- monious concurrence of fortuitous characters, probably existed in this country before the neolithic period, and was akin to or descended from the Belgian race of Furfooz. My Africanoid type, whose index of breadth, measured on the skull, would be about 74*5, may be a mere variety of the long-barrow race. But there are other skulls, or series of skulls, which have been measured and minutely described by careful observers, such as Huxley and Busk, and which are assigned to the neo- lithic period, but which depart considerably, in their general proportions, from the typical long-barrow cranium. Thus, the cave-skulls of Perth- y-Ghwareu,* in North Wales, seem to have a fairly arched contour, and an index of 76-5 ; while Laing and Huxley's Caithness skulls yield an index of 75-1.! I doubt whether quite sufficient attention has been paid to the greater breadth of these Perth-y-Chwareu and other cave skulls, as compared to those found in long and chambered barrows. After all, though the difference between an index in breadth of 71^ and 76^ is a mere trifle when we are dealing with individuals, it is of some importance when it represents the averages of two sets of skulls. We may talk of these Perth-y-Chwareu men as long-heads ; but their heads were, to say the least of it, quite as broad as those of modern Welshmen, and rather broader than those of our modern West of England folk, or of the old Anglo-Saxons. Either they were a different race to the neolithic folk of the long barrows, or we need to enquire into the causes of the difference of the twain. When no less an anatomist than Johannes Ranke starts afresh the doctrine that form of brain may depend on ossification of sutures rather than the opposite, it may be time to reconsider the doctrine of Barnard Davis, that the skulls of the long-barrows, or at least the * Boyd Dawkins's Cave Hunting, pp. 168-171. f Prehistoric Remains of Caithness, p. 161.

14 THE RACES OF BRITAIN.

extreme examples of narrowness among them, owed their form to premature synostosis, however the liability to such a process might have arisen. The so-called " river-bed " skulls, again, have for the most part a rather broad form, with a general resemblance to late or modern types, Irish or English, according to their place of origin ; but there is, I apprehend, no particular reason for attributing to these the great antiquity sometimes claimed for them.*

It is almost certain that no considerable body of people now exists in any part of Europe who would yield so low an index — that is, who have skulls so long and narrow — as the occupants of our long barrows. Yet, at the age when dolmens and chambered tumuli began to be built, there is reason to think that extreme dolichocephaly was the prevailing type throughout most of the west and north-west of Europe, from the Baltic to the Straits of Gibraltar.

French anthropologists think there were two palaeolithic races of long-heads in their country — those of Canstadt and of Cro-Magnon — before the neolithic long-heads, the first constructors of dolmens, came in from the north-east. If so, it is likely enough that the third race was a blond one, identical with the Tamahu, who are supposed to have imported dolmen-building and the fair complexion into North Africa, and who are portrayed with blue eyes and light hair upon the Egyptian monuments of 1500 b.c. Or, if not, the Tamahu may have been a subsequent wave of long-heads, but the first of the succession of blond northern conquerors, and may have taken up the dolmen type of civilisation, so to speak, as in the historical period the Normans accepted the civilisation of the conquered Frenchmen of Neustria.

The division of the European long-heads is at present sufficiently well marked into a northern blond race or races, extending from Flanders far to the north and east, around the North Sea and the Baltic, and a southern brown race, in Spain, Sardinia, Southern Italy, &c, which extends into Northern Africa. Cross exceptions in colour occur, however, not only in individuals, where they might arise from " sporting," but in considerable masses of population. Thus, certain Berber tribes — especially, it is said, some who dwell in districts where dolmens are very common — contain a large proportion of blonds ; while, on the other hand, we have in the western parts of the British Isles, as I shall hereafter show, a notable concurrence of long heads and dark hair; and M. Hamy finds in the dark-haired people of Dalecarlia descendants of the southern long-heads. f These exceptions are sufficient to prevent us from too hastily accepting the notion, commonly held ever since the days of Galen and of Celsus, that the colour of the hair

* Dr. Henry Bird, on the strength of observations on certain small, ill-developed skulls from small "tump" barrows, mostly in Gloucestershire, believes in another race prior to the long-barrow men. But the evidence is insufficient.

t The only Dalecarlian I have ever seen had gray eyes and rather dark hair and complexion ; his cephalic index was only 734, which would be 714 in the skull.

THE RACES OF BRITAIN. 15

depends simply on temperature and latitude; yet, though they indicate that hereditary influence overbears every other factor in determining the distribution of pigment, they must not be taken as proving that no other factor has any power whatever. If, on the one hand, we find that upwards of 3,000 years of Africa have not turned the descendants of the Tamahu into brunettes, we acknowledge, on the other, that their colour was probably developed and fixed for them in some original home in the temperate and comparatively sunless north. If it be true that a dark- haired type is even now flourishing in Dalecarlia, it probably found its way thither by following from the south the gradual northward migration of the reindeer. The history of the British Isles is that of an irregular or intermittent current of invasion from the neighbouring continent — invasion of ideas, of customs, and of arts, even more than of human beings. Anthropologically, Britain has been always a stage farther back in development than the Continent. Thus, in France more than one type of broad-headed men already existed before the building of dolmens was conceived or learned ; but it was the pure long- headed race who established the practice in England, though it is not impossible that the race of Furfooz may have existed among them in a state of serfdom.

The comparative date of interments in cists is usually very uncertain : this mode of burial probably preceded, and was contemporary with, the erection of long barrows and galleried tumuli, and certainly it survived them. But we may say with some confidence that most kinds of early round barrows came into vogue with the introduction of bronze, and that bronze and the bronze culture were brought into Britain by a broad- headed race.

The average index of breadth of 80 skulls whose measurements are given in the Crania Britannica, and which were not found in long barrows or chambered tumuli, is as high as 82 ; that of height, according to Dr. Davis's system, being about 76, which would probably be over 70 on the basio-bregmatic plan. Among the 80 are probably a considerable number not belonging to the bronze period ; and if these could be identified and subtracted (which is impossible), the breadth might have been greater. On the other hand, Dr. Davis measured the length of skulls, not from the glabella, or most prominent point between the superciliary ridges, but from the plain of the forehead, nearly an inch higher, in order to avoid the frontal sinuses, and approach more nearly to the length of the brain ; and this procedure, in a frowning, beetle-browed head, such as were those of most specimens of the bronze race, would somewhat increase the relative breadth. Thus, Davis and Virchow, measuring nearly the same series of stone-age skulls at Copenhagen, got average indexes, the former of 78, the latter of 77*3. Moreover, there appeared to Barnard Davis to be a certain degree of parieto occipital flattening in most of the round-barrow crania, the result, probably, of laying the infant's head on a cradle-board or other hard

1 6 THE RACES OF BRITAIN.

sul >st;mce. ::: On the whole, however, we cannot be far wrong in describing the British skulls of the bronze period as distinctly brachy- cephalic ; and this seems to have been the case in Scotland I as well as in England. Whether it was so in Ireland also we have not material for forming a judgment ; but it is probable that the bronze race did not settle numerously in Ireland, though MacFirbis's traditional account of the Tuatha de Danaan, the large, fair, vengeful race, skilled in music and in magical arts {i.e., bringing with them a higher civilisation), may, perhaps, point to them.

Whence came this race? Its strong resemblance to a type which abounds in the stone-age tombs of Denmark has struck several observers,^ the principal points of resemblance being the rounded form, the great parietal breadth, the great development of the superciliary ridges, and the prominent nose and chin. Except that the parietal breadth is moderate, these characters all occur also in the Sion type of His and Rutimeyer, the fathers of Swiss anthropology, who ascribe it to the Helvetii ; and there are skulls which the Swiss savans refer to a mixed type — simply, as it seems to me, on account of their greater breadth — which bear a yet closer resemblance.? The same characters appear in the modern Walloons, the descendants of the Belgse, and. with some attenuation — due, it may be, to admixture with the generally long-headed aborigines and Germanic conquerors — radiate from the Ardennes into the north-east of France, to enter into the composition of the Kimric race of Edwards and Broca. Virchow remarks that the Danish skulls of the iron period, which are very long, gave him the impression of belonging to a different race from those of the stone period : " Sie machen den Eindruck, als seien sie einem anderen Volke angehorig ;" whereas he evidently thinks the Borreby type is represented to a considerable extent in the present population of Zealand, which, so far as we are aware, possesses an intermediate index of head-breadth. I myself found one of 8o-6, which would be 78 or 79 in the skull, in 14 Isle-Danes.*

Looking at these facts in the light of the statements of the classical authors respecting the Cimbri, their original location in or about Jutland, and their south-westerly movement into Belgic Gaul, one is disposed to think the Borreby skulls may have belonged to a race, if not identical, yet nearly allied to the Cimbri, which may have been partly subdued,

* For my own part, I am disposed to think too little importance is attached to the mode of nursing infants, as an exaggerating cause of brachycephaly. I believe it to be operative in this way in some parts of Germany; e.g., Nassau.

t D. Wilson, Archaol. and Prehist. Annals of Scotland, pp. 168-171.

} Especially Rolleston, British Barrows, p. 631, &c. He remarks also (p. 712) that the frequent occurrence of amber in round barrows is an argument for the Baltic origin of the tenants.

§ See further on, in the chapter on Switzerland, a diagrammatic comparison of skulls of these three types.

|| "Die Altnordischen Schaedel zu Kopenhagen." Archiv fur Anthropologic.

• Heddoe On the Hcadform of the Danes, "Anthrop. Memoirs." See also British Assoc. "Report on Facial Characteristics;" respecting the prominence of the super- ciliary ridges in some of the modern Danes.

THE RACES OF BRITAIN. 1 7

partly expelled by a long-headed race of conquerors, Danish or Anglian, and which may have found its way across the Rhineland and Northern Gaul, acquiring the bronze civilisation on the way, into the British Islands. For other kinds of evidence, for or against this view, I must refer to the next chapter.

Whencesoever they came, the men of the British bronze race were richly endowed, physically. They were, as a rule, tall and stalwart ; their brains were large ; and their features, if somewhat harsh and coarse, must have been manly and even commanding. The chieftain of Gristhorpe, whose remains are in the Museum of York, must have looked a true king of men, with his athletic frame, his broad forehead, beetling brows, strong jaws, and aquiline profile.

What has become of them ? Perhaps they were never more than a dominant caste. At all events, continued admixture with other types, in the course of upwards of two thousand years, might well tone down their more salient characters, except in a moderate proportion of instances, in which atavic heredity may have preserved the type. Mr. Park Harrison, who has paid much attention to the point, finds that in those individuals who retain it in its greatest purity, the lips are thin and straight, the ear long and pear-shaped, the eyes bluish-gray, and the hair light ; and I am disposed to agree with him, though I think the colour less constant than the form. Among the men whom I have measured in the West of England and in Wales, 82 had heads with a breadth index of more than 80. The colours of their hair and eyes are given below, reduced to percentages ; and it will be seen that they are very much lighter, generally speaking, than those of the rest of the local population :

LIGHT

EYES.

NEUTRAL EYES.

DARK

EYES.

0 c

d

a

d

x 3

< X

T3

0 u

tab

O

-0

'3

0 u

u ni

"c3 •SP 0

ed. air. row

U cd

sb

0

,_, tap

tt,

PQ

Q

2

H

&

h

pq

A

Z H

K h PQ

Q

2

H

Z

4'3

20

262

128

—

633

—

24

3'6

97

12 16

•6 6 61

103

18

205

10 9

It has been noted that the great development of the brows, and the transverse furrow on the forehead above them,^are shared by this type with the Australians and some other savage races, as well as by the ancient Canstadt race, who have even been thought to retain in these points a Simian characteristic. Ranke and Kollmann say well, however, that points of likeness to the anthropoid apes are distributed variously among the different races of mankind, but that none of them can be taken in themselves to imply intellectual or moral inferiority. King Robert Bruce's skull was of the Canstadt type, and Savonarola's approached it. Certain it is that the British bronze type is found frequently — I should say with disproportionate frequency — among our best as well as our ablest and strongest men.

There is no distinct craniological or sepulchral evidence of the settlement of any subsequent race in Britain before the advent of the

10 THE RACES OF BRITAIN.

Romans. Historically, indeed, we know that immigration from Gaul continued after iron had begun to take the place of bronze. What manner of men the immigrants probably were I shall discuss in the next chapter, on other kinds of evidence. Meantime, I may say that the people whose land the Romans invaded, whatever their chiefs may have been, were not themselves, in the mass, of the bronze type. The inter- ments which we call Romano-British vary considerably; but more often than not, I think, they exhibit a type which Professor Wilson and I used to call the Celtic, and which he ascribed to a race of invaders posterior in date to the bronze folk. The name Celtic is better avoided for several reasons, and chiefly because our late master, Paul Broca, limited it, on grounds which he thought sufficient, to the race of men that predomi- nates in the old Celtic Gaul, from Bretagne to Savoy, whose short, thick-set figures and large, broad heads and faces are entirely different from the characters of the people in question. Wilson gave the name of " pear-shaped " to the skulls of these supposed Celtic invaders : the term was not, perhaps, very happy; nor is that of " coffin-shaped," which I suggested, much better. In fine, these skulls are intermediate in length and breadth between the long-barrow and the round-barrow forms : they have the prominent occiput of the former, with some degree of the parietal dilatation of the latter, and a long, flattened temporal region, gradually widening out to the point of maximum breadth, which is generally rather far back. This character belongs to neither of the other types, but seems to me a probable result of their partial fusion. If the existence and prevalence of this type were not brought about in this way, it must either have been the direct descendant of the river-bed and cave men, or have been imported by very numerous invaders from Belgic Gaul. Its resemblance to some forms that occur in Scandinavia, and in the Reihengraber of South Germany, is commented on by Huxley (Prehis. Remains of Caithness), and certainly the general proportions of all three are very much alike ; so much so that they might be difficult to distinguish with certainty, especially if the lower jaw were not present. Still, the British type usually differs in some details from the Scandi- navian : the temporal flatness, particularly, is more marked ; the forehead is flatter and squarer ; and the eye deeper set, not to speak of differences in colour. The Anglo-Saxon type inclines rather more to be elliptic in the norma verticalis (i.e., when viewed from above) ; the orbits are more rounded, the chin rounder, the jaw broader, the brows more arched and less prominent. Moreover, the general distribution of the British type through the three kingdoms tells strongly, though not absolutely, y against its being a late importation. The only apparent loop-hole of escape for those who think so seems to me to be the following : What if the Milesians, a race, by hypothesis, of Gaelic-speaking conquerors from Spain, perhaps somewhat darkened as to hair by Iberian admixture — what if they overran Ireland in force and naturalised their type there, while long-headed Belgae did the same thing in England and Wales ?

CHAPTER III.

Britain before tosar anfc Claubius.

WE may now proceed to discuss the ethnological position of Britain, especially its southern part, when the invasion of Caesar brought it within the domain of history. Caesar was informed, and believed, that the interior of the island was inhabited by an aboriginal race, and that those tribes who inhabited the maritime parts had passed over the Channel, and had mostly retained the names of the States from which they had proceeded. The lists of the tribes of the south and east of England, which we find in the works of the classical geographers, appear to corroborate this last statement ; at the same time, it does not exactly follow from such similarity of name that the Morini and Attrebates of Britain, for example, were identical with the Morini of Flanders and the Attrebates of Artois. Derivations for these names, and -those of most of the other tribes, may easily be found in the Welsh language, and were probably significant in the Belgic ; in which case similarity of position, rather than derivation of blood, may have been the cause of identity of name ; and the Morini of Dorset, to return to our example, may have been so called because, equally with the Morini of Belgium, they were seated on the coast, and not from being a colony of that tribe.

Tacitus affirms that the language of the Britons was " haud multum diversus " from that of the Gauls ; and we have the authority of Strabo that the dialects of Belgic and Celtic Gaul were very similar. So little curiosity had the Romans in general about the languages of barbarians, that we are not entitled to conclude, from the silence of Tacitus and the geographers, that there were no great diversities of the kind in that part of Britain known to them ; still, Tacitus did found speculations as to the diverse origin of the tribes on their physical aspect, and did not altogether ignore the evidence of language in other cases." His doing thus, by the way, adds to the probability that in his day the Silurians had ceased to speak an Iberian tongue, if they ever had so spoken ; for if it had not been so, Tacitus might have used their language, as well as their curly hair and swarthy countenances, to sustain his theory of their Spanish connection.

Again, the words acknowledged to be old Gallican (such as petor- ritum) are generally allowed to be explicable by the modern Kymric ; * See the Germania, with reference to the CEstisei and Gothini.

3*

20 THE RACES OF BRITAIN.

and there is a great resemblance between the Kymric and the Gallic local names, though both in Gaul and in Britain there are linguistic indications of the presence of Gaels." The names found on British coins — Boduoc, Commius, Dubnovellaunus, Tasciovanus, &c. — may be, in some cases certainly are, both Gallic and Welsh ; and the same may be said of the names of the chieftains of Kent, mentioned by Caesar; but none of them seem to be Germanic. The religious institutions also were alike on both sides of the Channel ; so, on the whole, were the arms and ornaments : the tore equally adorned the necks of the Boian warriors, of Boadicea, and of the sons of Llywarch Hen, and of the 363 Kymric chiefs who marched in insolent confidence to the fight of Cattraeth ; though the British cetra or target may have been Iberian or Gaelic, not Gallic. f

It is strange, in the face of all this evidence, inferential and circum- stantial though it may be, that some weighty authors have favoured the idea of the Germanic character of the Belgae as a people. Even Palgrave gave it a doubtful adhesion ; and Latham does not always and altogether oppose it. The grounds alleged in its support appear to me extremely weak.

They are : firstly, the difficulty of accounting otherwise for the rapid Saxonization of England ; secondly, the nearer approach of the Welsh than of the Gaelic to the German tongue ; thirdly, certain statements of Caesar and others as to the German origin of some of the continental Belgae ; and, fourthly, the possibility of the Galatians of Asia having been Germans.

The consideration of the first of these objections may be deferred to to my fifth chapter. The second hardly touches the question : the ancestors of the modern Welsh, in blood and language, were the Silures, Demetes, Ordovices, Ottadini, &c. ; and these are scarcely suspected of having belonged to the recent Belgic immigration. Very early Belgian colonies, to which the Fir-bolg may possibly be ascribed, might be Keltic or Ibero- Keltic, but must have occurred before anyone dates the appearance of Germans to the west of the Rhine.

As for the third objection, that there were already, in Caesar's day, some truly Teutonic tribes on the left bank of the Rhine, is probable enough, though their presence there was pretty surely recent.

Though Caesar had been told by the Celtic Gauls that most of the Belgae were of German descent, more precise information seems not to

* Guest and Peacock. The former thought the Belgic tongue was Gaelic rather than Kymric ; and Leo, quoted by Taylor, partially supports him. Taylor, Words and Places. Davies, Celtic Researches. But most of their examples are doubtful. Basil Jones, Vestiges of the Gael. But the local names given by him may have been of later introduction, though probably not.

t Cran. Brit. Much confusion arises from the likeness, in their modern forms, of the words Gallic and Gaelic. Some French writers reason from their supposed identity.

THE RACES OF BRITAIN. 21

have confirmed this statement ; for he subsequently specifies ':: four small tribes — the Condrusi, Eburones, Cceresi, and Poemani — as Germans, apparently to the exclusion of the other and greater tribes. And even the Eburones, by the way — if we may draw an inference from the name of their great war-chief, Ambiorix — were but doubtful Teutons. Tacitus, who evidently aimed at accuracy in the matter, allows only the Tnbocci, Nemetes, and Vangiones, all close to the Middle Rhine, in the modern Elsass, to be " haud dubie Germanos;" adding, that the Treviri and Nervii affected a Germanic origin because it gratified their vanity. The modern population of Hainault, the ancient territory of the Nervii, is Walloon in language and in physical type,! notwithstanding its having been occupied by the Franks ; and therefore is probably not of German origin. To the Treviri I will return presently, as their case is important. The Gallicism of the great tribes of the Remi, Bellovaci, Suessiones, and Attrebates has hardly ever been doubted ; and the name of the last- mentioned, and the fact that Divitiacus of Soissons made conquests in Britain, lead us to derive some, at least, of the British Belgae from this quarter.

Gallic skulls from the region of the Bellovaci, so far as I have seen, differ considerably in general from those of the Merovingian Franks, but much resemble certain skull-types common in the West of England ; and these skull-types are of very old date in Belgium and Northern France, occurring in remains from the sepulchral caves of Nogent, as well as at later periods. And the figures of the provincials on the monument of Jovinus, in the Cathedral of Reims, beautifully discriminated in type as they are from the Roman Governor, are equally and conspicuously unlike the modern Germans in feature ; while, notwithstanding the great influx of German Franks that must have poured into the Belgic provinces in times sub- sequent to that of Jovinus, the modern inhabitants of Reims are mostly quite unlike Germans in complexion, and exhibit, equally with the monumental figures, the tall frames, square foreheads, and long, sharply- drawn features, which constitute William Edwards's Kymric type.| The same complexion and features continue to prevail, with little modifica- tion, as one journeys north-eastwards through the Ardennes, as far as Liege and Verviers (where many of the Walloon inhabitants reminded me strongly of Cornishmen) ; but they suddenly change, and give way to German characteristics, when we cross the frontier of the Walloon dialect in the direction of Mechlin or Aachen. Seldom can we see an ethnological line so distinctly drawn as that between the Walloons and the Flemings, though for hundreds of years they have been under the same government. My own observations show the facts pretty clearly, the indices of nigrescence for the Flemish-speaking country varying from a minus number up to + 26, and those for the Walloon districts

* Bell. Gall., ii.4. t Vanderkindere, Recherches, &c.

\ See the frontispiece for a comparison of the Belgic with the old Frankish type ; the latter still visible in the peasantry of the Lower Rhine.

22 THE RACES OF BRITAIN.

from 28 to 61 ; but the beautifully displayed statistics of Professor Vanderkindere are equally distinct, and far more extensive and weighty. Of the 26 arrondissements into which Belgium is divided, in 14 of which Flemish is spoken, and in 12 Walloon, every one of the 14 shows more blonds and fewer brunettes than any one of the 12. At the same time, M. Houze demonstrates a decided difference in the breadth of the head between the Flemings and the Walloons. It is difficult not to be con- vinced that we have here to do with a real ethnic frontier, and that we have on the north a Germanic, on the south an old Belgic type. I may be reminded that the old Reihengraber type, to which the Franks and Alemanni are supposed to have belonged, was long-faced as well as long- headed. So it was, to some extent, but not so remarkably and generally as the Kymro-Belgic form I have been describing. Moreover, it both was and is usually combined with fair hair ; whereas we have seen that the Walloon type includes comparatively dark hair in the present day ; and in all probability, though the chiefs, the true Galatae, were fair, the mass of the old Belgae was of old something like what it is now.

The modern Treves is now within the limits of the German tongue. I should say that the German physical type also prevails there, though not so as to overpower all others ; but the city, lying as it does on the eastern side of the principal masses of frontier hill-country, may probably have been Germanised during the great westward movement of the Ripuarian Franks in the fifth century.

But the main point of interest about the city, for us, is the fact, avouched by Jerome, that the Asiatic Galatians spoke the language of Treves. Now nobody, until lately, ever dreamed of denying the Gallicism of the Galatians. Niebuhr, having made the Treviri Germans in spite of Tacitus, evaded the difficulty by the futile explanation that Jerome's Galatians were not Galatians at all, but Goths from Thrace. In this, I believe, no one has followed him ; and Latham has rather severely characterised the criticism in his Germania. But others have made the Galatians German, on the ground that the names of the two chiefs who led them across the Hellespont, Leonarius and Lutarius, are very like the German Leonhard and Lothar (as they certainly are). To this, however, a sufficient answer seems to be, that of the three principal Galatian tribes, the Tectosages undoubtedly, and the Trocmi and Tolistoboii apparently, bore Celtic names ; and that the same may be said of their subordinate tribes, the Tosiopi and Teutobodiaci ; of most of their chiefs, e.g., Brennus, Bitoetus, Deiotarus, Brogitarus, Bogodia- torus ; and of some of their towns, e.g., Eccobriga and Tolosochorion. Taylor {Words and Places) gives a list of local names in Galatia, con- taining some others that might support my argument ; but he seems to be in error when he quotes several which embrace the root Mag or Myg to show the Gaelicism of the Galatians ; some at least of these words, such as Magnesia, having existed long prior to the arrival of the Galatians in Asia, though they may possibly date from that of the Kimmerians.

THE RACES OF BRITAIN. 23

In the Triads the Belgae seem to be spoken of as " the refuge-taking Men of Galedin;" and the part of South Britain extending between Kent and Somerset — i.e., precisely the part which the Belgae did certainly occupy — is called Arlechwedd Galedin in a list of the divisions of Britain. :;: To return to the physical characteristics of the Belgae, we know that they appeared to the swarthy Romans to be a tall and rather fair race ; but if they had been generally so tall and fair as the Germans, Caligula would have had no need to choose the tallest of them, and dye their hair, when he wished to make them pass for German captives.

On the whole, I have little difficulty in concluding that the Belgae — though there was in some of their tribes, and perhaps in the noble and military caste throughout, some infusion of German blood ; and though former waves of long-headed and light-complexioned warriors, Kymric or Galatic, may have rolled over their country and mixed blood with them in earlier ages — were a Celtic-speaking and, to some extent in blood and physique, a Celtic or Celtiberian people in Gaul, Asia, and Britain.

There remain, however, still some arguments from other quarters in favour of the presence of Germans in Britain at this early period.

Of these, one is derived from the statements in the Triads, and else- where in the old Welsh literature, respecting the Coranied ; another from the mention by Ptolemy of Petuaria and a Gabrantuicorum Sinus, in the country of the Parish, north of the Humber. The Coranied are said in the Triads to have been one of the races who came into Britain in a hostile manner, and to have come from the land of Pwyl, which is by some interpreted to mean the marshy lands (pools) at the mouths of the Rhine, but by Lord Strangford, who discredited the whole story, to have been some mediaeval Welshman's notion of Poland. They are also said to have treacherously made common cause with the Saxons in their invasions. In the tale of Lludd and Llewelys, in the Mabinogion, are some fantastic stories about them, in which they are represented as a race of magicians (reminding one somewhat of the Tuatha de Danaan of Ireland) who were destroyed by King Lludd. According to E. Davies,t one of this race was called Cawr or Cor, a giant ; in the plural, Coried or Corion. All this sounds very mythical ; still, to those who find in the Jotuns of Scandinavian mythology merely the primitive Finnish inhabitants, it will present little difficulty. The Coranied are identified with the Coritavi or Coritani of the Romans, from the similarity of the first syllable in each word, from a statement that the Coranied settled about the Humber, and from the name of Ratis Corion having been applied to Leicester, seemingly the chief town of the Coritavi. The only grounds for making the Coranied and Coritavi (allowing them to be the same) Germans are, their siding with the Saxons, and having a Latin name ending in avi, like the undoubtedly Germanic tribes of the Batavi and Chamavi. I entirely disagree with this view, for the following reasons :

* Iolo MSS., p. 477. t Celtic Researches, p. 200.

24 THE RACES OF BRITAIN.

They are supposed to have occupied the counties of Lincoln, Not- tingham, Derby, Leicester, Rutland, and part of Northamptonshire ; and in these counties I can find no Roman station whose name appears to be Teutonic ; while the important town of Margidunum, near South- well, in Nottinghamshire, bears a name almost certainly Celtic, and Ratis Corion does the same ; and Nottingham would seem to have remained Celtic long enough for its Welsh name not to have been altogether forgotten even in the time of Alfred ; for Asser says it was called in Welsh, Tigguocobauc. Again, if the Coritavi were Germans, and were overlaid by successive strata of Angles and Danes, one may reasonably expect to find the Teutonic physical type prevalent over their whole area to a degree not found elsewhere in Britain. Now, in the northern part of the Coritanian area it is really very prevalent, but in the southern (Leicestershire and Northamptonshire) there is, if I may judge by the colours of the hair and eyes, a strong non-Teutonic element. The following table shows a great difference between Lincoln and Leicester, Nottingham and Northampton, in these respects, there being a much larger proportion of dark hair in the two more southern towns :

<~ EYES LIGHT. EYES NEUTRAL. EYES DARK. ,_ «

«-T3 ° O S

J> ?« a 6 d %%

6 55 o<_- c S m . . ■ > * . .,.■$* . %9

££ i~1 -g g 3 .SP "g -s E 3 .2? "S '3 8 S g =.2?

Lincoln 500 2 15 35^2 85 2 '4 12 5'2 4^4 -2 -g 1*3 8-i i5-8 17 iz'3

Leicester 540 3 13-9 266 yi -i -8 r6 66 6'8 '2 -4 13 y8 ig'g 37 20-8

Nottingham 700 37 153 24'6 99 -3 1 18 61 5Pi 'i 11 14 9 182 22 i4'i

Northampton ... 300 3 98 356 13-3 -8 -3 7 35 4-5 — -3 -5 5-3 187 3-8 3fl

Professor Phillips, than whom no ethnologist was a keener observer, once visited Leicestershire, with the expectation of finding a strongly- marked Scandinavian type predominant there ; but he was surprised to find a dark-haired type, which he supposed to be Celtic, equally prevalent. This may easily be accounted for, and that without treating the traditions about the Coranied as altogether spurious, as Lord Strangford thought them, if we suppose the Coritavi to have been a colon)' of Celtic Belga? ; but, unless we throw aside the evidence of physical type, we can hardly conceive how the)' can possibly have been Germans. Moreover, the silence of Tacitus respecting any suspicion of there being Germans in the island, except the Caledonians, is of weight on the same side.

The names Petuaria and Gabrantuicorum Sinus do certainly, at the first blush, look rather Teutonic than Celtic ; but the suggestion of Mr. Isaac Taylor, that the Parisi or Parish may have been Frisians, does not commend itself to my mind. The name of the Frisians was too well known to have been thus distorted ; and Ptolemy would probably have called them (frpiaaovc-t or (fipiaioi, as Procopius did subsequently. If Whitaker's etymology"'" be trustworthy in this instance, as I incline to think, the

* History of Manchester, i., 45.

THE RACES OF BRITAIN. 25

Parisi were simply Paruis, herdsmen, and the Gabrantic goatherds. Thus fades away, bit by bit, all the evidence in favour of the presence of Germans in the southern part of Great Britain before the Roman period.

Tacitus, whose ideas are always entitled to consideration, thought the Caledonians were of German origin ; but the only grounds he assigned for this opinion were their huge limbs and fiery-red hair (vutila; coma, magni artus). If their language had been German, he would probably have known and mentioned the fact, as he mentioned the Gallic language of the Gothini and the CEstiaei. Modern philologists do not, I believe, pay any respect to the notion of Jamieson, that Teutonic names existed in the north-east of Scotland prior to the arrival there of the Angles and the Norsemen. The red hair and large limbs are still prevalent in Athol and Marr and Badenoch (see 51, 52, 53, &c, in my Scottish table) ; but all we can say of them is that they point to an origin from the northern rather than from the Mediterranean long-headed races. The Caledonians might have come over from Denmark, and yet borne their Celtic name ; but to one who looks at them from the point of view of the physical anthropologist, it may seem more likely that they were a Gaelic or a Pictish tribe, with a strong dash of the athletic broad-headed element.

In Ireland the name of the Cauci suggests a Frisian colony ; but I know of nothing else to favour the notion. The Irish traditions indicate that there, as in Great Britain and in France, successive swarms of fair- haired invaders overlay the dark-complexioned aborigines ; but that any of them were German, in the strict sense, is improbable. It may be worthy of remark that the inhabitants of the Aran Isles, in Galway Bay, are reputed to be descended from the Firbolg, who were masters of Ireland before the advent of the Tuatha-de-Danaan, and that they have nearly the same long-featured, long-headed type already spoken of as common in the Belgic region of Northern France. In this connexion, the opinion of Dr. Guest, that the Belgic dialect belonged rather to the Gaelic than to the Kymric branches of the Celtic tongues, becomes very suggestive.

There remain other questions, similar to that of the presence of Germans in England before Caesar, which present still greater difficulties; such as the following : Were there considerable remains of the Gael among the Cymry and the Lloegrians of England and Wales ? Was there a notable proportion of Iberian blood among the Britons, especially the Silures and the people of Cornwall ? Who were the original Lloegrians ? and had they any connexion with the Ligurians beyond the similarity of their names ? Did the Phoenicians really trade directly with Cornwall ? and if so, did they leave there any traces of their blood ?

Until of late years, almost all we had to show for our belief in the existence of an Iberian substratum in our population were, the con- jecture of Tacitus respecting the Silures ; the length of head in the long-barrow people and some other neolithic men ; the resemblance

26 THE RACES OF BRITAIN.

between the Welsh cave-men and Busk's Gibraltar skulls ; and the sup- posed greater frequency of dark hair, especially in the west, than could otherwise be well accounted for. I hope to be able, in a later portion of this book, considerably to define and strengthen the evidence of physical characteristics. In the meantime, I will say that dark eyes and dark hair, often curly, are still very frequent in Siluria, but that the dark colours are not much less so in Dyfed and Gwynedd, the other ancient divisions of Wales. Dark and even black hair is abundant also in Cornwall and Devon, and in those parts of Scotland and the North of England where Kymric blood may well be supposed to remain in large proportion, such as Upper Galloway, Strathaven, and Allendale. And Strabo says that in his time the Britons were somewhat (taller and) darker-haired than the Gauls ; so that we must suppose that rather dark hair was frequent among the ancient Britons in general, though perhaps especially so among the Silures. I am prepared to admit that a physiognomy strikingly Iberian (or Basque-like, at least) is commoner in South Wales than in any other part of Great Britain. Many photo- graphs of Basques, and some of Bearnese, are recognised, both by myself and by an observant Welsh anthropologist to whom I have submitted them, as being in no respect different from some of the ordinary types of feature in South Wales.*

Anthropologists have long been awaiting the appearance of some philologist fully qualified to determine the important problem whether there be really an Euskarian or Iberian element in the Kymric languages, or, if so, whether it be equally or more potent in the Gaelic and Erse. The existence of such an element had been boldly asserted, and super- ciliously denied or ignored, until recently Professor Rhys has answered our call with the assurance that the element which physical phenomena has led us to look for does really exist ; that it is to be found in Gaelic rather than in Kymric, and in Pictish rather than in Gaelic ; and that the Iberian symptoms among the Silures must be accounted for by their having been, in part at least, Gaelic before they became Kymric in language.

Professor Rhys's opinion is clear and consistent, and may be recon- ciled with physical facts better than any other hypothesis on the subject, as I shall subsequently demonstrate. The well-reasoned affirmation of Skene, that the Pictish language was, phonetically, intermediate between Gaelic and Kymric, is the principal obstacle in its way, and that is not, perhaps, insurmountable. The Pictish of Skene, and even of Bede, may have been the Celtic language in the form which it took in the mouth of the Picts ; and it is noteworthy that the Picts had been brought intimately into contact with tribes some of which spoke Brythonic, and some Gaelic dialects.

The evidence in favour of the importance of Phoenician intercourse with Britain, and of the introduction of at least some little Punic blood * Greatheed made the same remark, Archaologia, xvi.

THE RACES OF BRITAIN. 27

into Cornwall, may be found in the chapter of the Crania Britannica on the " Historical Ethnology of Britain." The statements of Diodorus and others certainly lead one to suppose that the intercourse had had a considerable effect ; for the grave, black-robed people, described as inhabiting Cornwall, nowise remind us of the British dress, aspect, or character. Traffic with the Phoenicians must have involved intercourse with Spain, but to what extent is quite inscrutable. The maritime position of Cornwall, and its many good harbours, have led to the intro- duction of a variety of ethnological types and elements, of which the Semitic is one : the Iberian was pretty surely there previously. That the long-barrow race can have been a ruling caste of Phoenicians does not seem possible.

The presence of Gaelic-speaking folk in Great Britain has already been touched upon. The subject has been chiefly in the hands of the philologists, and the evidence upon it has been gradually growing stronger. There are those who think the long-barrow race, if Iberian in blood, were Gaelic in speech ; and they adduce Gaelic etymologies for the names of places where chambered barrows and other megalithic monu- ments exist. ::: If the bronze race spoke Kymric, as is generally believed, and if the long-barrow people still spoke Iberian, it is a little difficult to identify the Gael in the craniological calendar.

The opinion of Lhuyd, that most of the river-names of Great Britain must have been bestowed by the Gaels, is not capable of complete proof, from the near relation of the two branches of the Celtic, but it appears likely enough. Some of them are, perhaps, older still, dating from the Iberian or pre-Gaelic period.! Bishop Basil Jones supplies, in a work which seems to me a model of dispassionate criticism,! a considerable amount of evidence of the presence of Gaels in the coast districts of North Wales and in Brecon at or soon after the time of the Roman invasion, and Professor Rhys a good deal more — weak in detail, and of a presumptive and inferential character, for the most part, but strong in the aggregate — to indicate that the greater part of Wales was more or less Gaelic even to the later period when the Kymric Britons from the north subjugated the whole of that country; and that Devon, on the evidence of the Ogham inscriptions, was still partly Gaelic during the Roman dominion.

The presence of Irish physical types in Wales may, of course, be accounted for by the common ethnological elements in the diverse branches of the Celtic-speaking folk, or by the later immigrations from Ireland ; but if I am correct in my belief, based on repeated observa- tions, that persons of thoroughly Gaelic aspect § are common in the

* Thus Dr. Henry Bird. t Hyde Clarke. \ Vestiges of the Gael.

§ That is, with dark brown hair, gray eyes, long heads, flat in the temporal and prominent in the upper occipital region ; with cheek-bones prominent rather than broad ; jaws often prominent, but somewhat narrow. Such persons are occasionally, but rarely, seen in other parts of the south and even of the east of England. Dr. Mackintosh found them plentiful in Dorset.

28

THE RACES OF BRITAIN.

Mendips and in Exmoor, while we know of no Irish immigration into Mendip during the historical period, and while these are precisely the districts into which a conquered race might flee for refuge, the fact furnishes a slight additional argument in favour of the views of Lhuyd and his followers.

As for the Lloegrwys, the view that identifies them with the Belgae is nowise favoured by the Triads, which speak of them as very near akin to the Cymry, and distinguish from both of them the men of Galedin, who were pretty certainly, and the Coranied, who, if anything, were probably, Belgic. The assertion that the Lloegrwys came from Gwasgwyn or Gascony would be in no way decisive, even if we could at all trust to it as anything but a mediaeval fiction ; for the Volcae Tectosages of Tolosa appear to have been the same people with the Tectosages of Galatia, and these latter, as we have seen, spoke Belgic ; so that Belgic tribes had penetrated pretty early as far as the Garonne. The temptation to identify them with the Ligurians is very strong ; but whatever the old Ligurians were, they differed considerably from the Cymry. As met with in north-western Italy, they seem to have been a very broad-headed race — brachiocephalic, indeed, to a degree very rarely met with in our islands, even in isolated cases." The theory that connects the Lloegrwys with the river Liger (Loire) is less open to objection. On the whole, we shall run least risk of violating probabilities based on the Triads, and on what little we know or surmise of their physical characteristics and language, if we suppose them to have par- taken somewhat of the blood of that great stock which, whether we call it Ligurian, or Celtic, or Arvernian, has survived the conquests and migra- tions of upwards of 20 centuries in the centre and south of France, and furnished the ancestry and the physical type, and perhaps the language too, of the swarthy, broad-headed people who prevail there to this day. The Brythons may have been derived from a more northern part of Gaul, where the tribe of the same name inhabited Picardy, and accord- ingly partaken more of the Belgic blood. The physical characteristics of the descendants of the Strathclyde Brythons are nearer to those of the Walloons than are those of the modern Welsh. The great stature of the men in some qf the vales of Upper Galloway, where Strathclydian blood is probably pretty well preserved, is very remarkable, and dis- tinguishes them from their relations in the Principality. The average stature of upwards of 70 men in the Glenkens district, measured by the Rev. George Murray, of Balmaclellan, exceeded 5 ft. iof in. (179 centi- meters) ; and in some other districts about the Border, where the old British blood is more or less mingled with Anglian, Norse, or Gaelic, it is little less remarkable.!

* Nicolucci, in 10 old skulls from Liguria, found an average length-index of 867. Calori has shown that brachycephalism prevails in Northern Italy generally. Language and head-form both indicate that their connexion was with the Southern Celts rather than with the Iberians.

t Beddoe On Stature and Bulk, pp. 32-43. Report of the Anthrometrical Committee of the British Association.

THE RACES OF BRITAIN. 29

To sum up this chapter, the natives of South Britain, at the time of the Roman conquest, probably consisted mainly of several strata, unequally distributed, of Celtic-speaking people, who in race and phy- sical type, however, partook more of the tall blond stock of Northern Europe than of the thick-set, broad-headed, dark stock which Broca has called Celtic, and which those who object to this attribution of that much-contested name may, if they like, denominate Arvernian. Some of these layers were Gaelic in speech, some Cymric ; they were both superposed on a foundation principally composed of the long-headed dark races of the Mediterranean stock, possibly mingled with the frag- ments of still more ancient races, Mongoliform or Allophylian. This foundation-layer was still very strong and coherent in Ireland and the north of Scotland, where the subsequent deposits were thinner, and in some parts wholly or partially absent. The most recent layers were Belgic, and may have contained some portion or colouring of Germanic blood ; but no Germans, recognisable as such by speech as well as person, had as yet entered Britain.

CHAPTER IV.

Gbe IRoman periofc.

TO what extent was the ethnological position of Britain modified during the Roman period ?

This question may be analysed into several, and made to read as follows : Was the destruction of the native Britons extensive ? What was the magnitude and character of the immigration brought about by the Romans ? Did the Gaels on the west, or the Saxons and kindred tribes on the east coast, effect any considerable settlements before the departure of the Romans ?

With the exception of the campaigns of Suetonius, it does not appear that the reduction of southern Britain to a province was attended with any extraordinary destruction of life. Agricola and some other generals who took part in it are described as men of excep- tional mildness. Still, the tender mercies of the Romans were cruel ; and when Tacitus put into the mouth of Galgacus that famous sentence, " Ubi solitudinem faciunt, pacem appellant," he must have felt that there were pretty good grounds for the charge he thus made against his countrymen. The Iceni and other tribes who joined in the revolt of Queen Boadicea may probably have been almost exterminated ; but the relations of the empire with the Germans at that period were not such as to justify the conjecture of some writers, that the vacant seats of the Iceni were filled up with settlers from the German coast, a conjecture that seems to me baseless and extravagant.

In process of time Britain was traversed by numerous great roads, and studded with Roman colonies and stations and villas, the remains of which are found generally, it is true, in the immediate neighbourhood of the roads, but sometimes at a considerable distance from them. In Gloucestershire, Somerset, Kent, Essex and Hampshire, and along the line of Hadrian's wall, the Roman or Romanised population must have been considerable; and even in the more remote, rugged, and unattrac-, tive parts of the island it is difficult to get more than a few miles away from some trace or other of Roman occupation. On the other hand, in situations favourable for the preservation of such remains — for example the Wiltshire downs — are found the vestiges of British hut- villages, whose date is fixed in the Roman period, and sometimes rather late in that period, by the evidence of coins disinterred.

THE RACES OF BRITAIN. 31

Different writers have taken very different views of the position of the native races under the Roman regime, views varying from that of Whitaker, who believed that most of the British tribes retained under the Roman sovereignty some remains of their autonomy (pretty much as Sikkim, for example, does under the British, notwithstanding the existence within its limits of important British stations like Darjiling), to those of Thomas Wright, who seemed to think that hardly any British blood was left except among the servile peasantry, the towns having been occupied exclusively by the people of mixed or diverse descent introduced by the Romans from every other part of the empire.

The population of Italy, already, at the beginning of the Imperial period, exhausted by the selfish policy and destructive wars of the Roman aristocracy, could hardly, even during an occupation of 350 years, have spared enough of their increase to re-people Britain, while they were equally or more intensely engaged in colonising Spain, Gaul, Africa and the Danubian countries. The Spaniards had almost an equal period allowed to them for the colonisation of their American dominions, where they had to do with nations who may be said, without any glaring error, to have borne to them in respect of civilisation and military character something like the same relation that the inhabitants of Western Europe did to the Romans. But except in Chili, where the climate is particularly congenial to European constitutions, and where the aboriginal population was but scanty, they did not succeed in making themselves the numerical majority ; and as soon as the stream of immigration was cut off, their numbers and importance, relatively to the aboriginal races, began to decline.

But the Roman system of transferring the military population of every subject territory, in small bodies, to various other parts of their dominions, where they could not help losing their own nationalities, and acquiring that of their masters, renders this analogy of very little value. The Romans may be said to have had at their command, as Romanising agents, as large a proportion of the population of their whole empire as they found it convenient to levy for military purposes. The Notitia Imperii shows us that bodies of Syrians, Cilicians, Spaniards, Moors, Thracians, Dalmatians, Frisians, &c, formed the military colo- nists of the stations in Britain ; and when even the emperors themselves were often not of Italian birth, and the most trusted officers and governors provincials or even barbarians, we have no reason to suppose that any notable proportion of genuine Roman blood found its way to this country. " No doubt," says Mr. Wright, " the colonists of these towns were accompanied or followed by their relations and friends ; and as evidently they were recruited from their own countries, they must have gone on increasing and strengthening themselves. They were all, however, obedient to Roman laws and institutions, used the Latin tongue, and had, indeed, become entirely Romanised." It is doubtful, however, whether a settlement of Moors, for example, could have even

32 THE RACES OF BRITAIN.

maintained its numbers without large and constant immigration. So extensive was the empire, that the natives of its southern and eastern borders could hardly have become fairly acclimatised to the cold and humid regions of the north-west. I shall return to this point presently ; it is of some importance in explaining the disappearance, in most cases, of all traces of the blood of these colonists, a disappearance which may also depend, in part, on the heterogeneous character of the Romanised population taken en masse, which would render its potency, in breeding, very inferior to that of a comparatively uniform and pure-bred race like the British rural population, among which it was dispersed. If we seek for light by investigating the effect of Roman occupation on language and physical type in other countries than our own, we find the evidence somewhat diverse and doubtful, yet not without interest and value. Thus we find that a comparatively short occupation sufficed to destroy the old language of Dacia, and to modify, in the opinion of some observers, the Dacian physical type by the introduction of an Italian one, which has not even yet been quite worked out ; that a much longer occupation replaced the Iberian tongues by a Romance language in the greater part, but not the whole, of Spain ; and that it was only just long enough to produce the same effect in Gaul, where a Celtic tongue sur- vived the Roman dominion, but had already received its death-blow, so that it languished and died out. In both the latter cases the physical type seems to have been a little, probably only a little, modified. The case of Armorica would be very valuable for our purpose, if we were not in the dark as to the extent and effects of the immigration into that country from Britain, from the time of Maximus (the Maxen Wledig of the Welsh) downwards. We know that the Veneti, the leading people of Armorica, had been ruthlessly extirpated as far as possible by Caesar, and that the whole country had been, like Britain, seamed with Roman roads and dotted with towns and villas. So far the cases are parallel. But a few hundred years afterwards, there having been no disturbing agencies of any great account in the meantime, except the immigration from the kindred and like-historied land of Britain, we find scarcely a trace left of the blood, language, polity or religion of the Romans ; and the chief enemy which Christianity has had to struggle against, and which it is said not to have succeeded in altogether eradicating even in our own day, is a Paganism attaching itself to the dolmens, the maenhirs, and the natural objects of the country, evidently the direct descendant of that which ruled before the introduction of the Roman ideas.

Very similar would probably have been the history of our own country but for the interference of the Saxons. In proportion as the central authority grew weak, the spirit of nationality among the Britons seems to have revived. About the second and third centuries of the Roman dominion we find, so far as I am aware, hardly a trace of a truly British name, owing, as I conceive, to the fact that every Briton of consideration, in public transactions, such as dedications, &c, used a

THE RACES OF BRITAIN. 33

Latin name in compliance with fashion, though he probably enough bore among his own people a purely Celtic one, pretty much as a Maori chief nowadays has very frequently two names, one of which is an English one, often borrowed from some friendly colonist.

About the year 408 we come upon a " Count," an officer of rank and talent, and the main pillar of the usurper Constantine's fortunes, who bore the name of Gerontius. Whether this was the Greek Yepomio*, or a translation of the Welsh Geraint, I cannot say ; but thereafter, with the exception of those supposed to be members of the Ambrosian family, we find none but Celtic names among the chiefs of the British people.

If any part of England, except perhaps Kent and Essex, was more thoroughly Romanised than the rest, it must have been the tract traversed by the wall of Hadrian. To the import of the disappearance of the names of almost all the towns on the wall I shall have occasion to revert in a succeeding chapter; it has, I think, in conjunction with similar facts in other quarters, an important bearing on the nature of the Saxon conquest. But I wish now to point out that in the only instance, out of 19, in which the Roman name has survived, that of Carlisle, the ancient Lugubalia, the Welsh Caer-luel, that name sounds very like a Welsh one latinised ; and that the only other, in which any relation at all between the old and the new name can be made out, is that of Magna, now Carvoran, where Carvoran seems to be a Celtic name signifying " the great castle," probably meant as a translation of Magna : from this I draw the inference that Magna had lapsed into barbarism and Britannicism before that region was conquered by the Angles.

If we extend this examination to England in general, we shall find that throughout the whole land the Roman local names that survive bear but a small proportion to the British ones ; and that where such Roman local names do still exist they are usually British ones latinised, e.g., Lincoln, Wroxeter, Cirencester, &c. Spinae, the modern Speen, is, however, a notable exception ; Pennocrucium (Penkridge) and Broca- vium (Brougham) perhaps only apparent ones. The Kentish instances I pass by for the present.

It does not seem possible to extract anything of value respecting this period from Geoffrey of Monmouth, nor yet from the fragments of uncertain date and origin contained in the Iolo MSS. If we could give any credit at all to them, they would add weight to the opinion of Whitaker, that British chiefs continued to hold a certain sway over their own compatriots, much as the chiefs of Arab tribes in Algeria are allowed to do at present by their French masters. There is indeed one set of traditional accounts which, though contradictory in themselves chronologically and genealogically, must surely contain some kernel of fact ; I mean those relating to the presence of Gael in Wales, and espe- cially in Gwynedd and the modern Breconshire, which I have already

4

34 THE RACES OF BRITAIN.

referred to. Professor Rhys points out that Serigi, the Irish chieftain said to have been expelled from North Wales, is probably Sitric the Norseman (Sigtryg with the silken beard ?) pre-dated several hundred years ; but though he distrusts much of the evidence of the presence of the Gaels in the third or fourth century, he does not doubt the fact. The real doubt is, whether the Gaels were Irish invaders, the Scoti mentioned as having raided in Wales in the time of King Niall-of-the- Nine- Hostages, or were the original occupants of the country. The latter theory is supported not only by the numerous arguments adduced by the Bishop in its favour, but by the extreme improbability that the Romans would have allowed the Irish Gael to acquire by violence settled possession of a large portion of one of their provinces.

This destruction of the Gaelic dominion in Wales by the northern Cymry, under the leadership of the sons of Cunedda, seems to have taken place somewhere about the time of the supposed invasion of Kent by Hengist ; but we may as well note here, by anticipation, the probable ethnological results. The Gael in Gwynedd were not extirpated ; and ages afterwards we hear of the fifteen royal tribes who were regarded as pure in blood, because uncontaminated by marriages with the descend- ants of the conquered Gwyddel (Gael). A large part of the population may have remained Gwyddelian in blood, and even in sympathies, for some generations ; else how should we account for the position, as national heroes, sorcerers though they be, which is held in the Mabi- nogion by the princes of that race, Gwydion ap Don and Math ap Mathonwy ? The account of the occupation of Garth Mathrin (Brecon) by the Gwyddel is tolerably clear, and contains no discrepancy or improbability ; and we are told that their descendants continued to dwell there in the historical period, intermixed with the other natives.

Meanwhile certain intertribal movements were taking place in Ireland and Scotland, of which we know very little with any approach to cer- tainty. The power of the Milesians, probably a Celtiberian race from Spain, who are described by Mac Firbis as traditionally "white of skin, brown of hair," may have been still growing at the expense of prior Celtic-speaking conquerors, and of the Cruithne or Picts, who still remained a separate nationality in parts of Ulster, if not elsewhere. Whether the early Oghams belonged to them or to some of the prior conquerors may be a little doubtful ; but I think Mr. Brash and General Pitt- Rivers have made out a strong case for their having been the work of a tribe who crossed over from Spain to Munster, and perhaps also to Devonshire. The Scots, a Milesian tribe, or at least under Milesian leadership, were beginning to make settlements on the western side of the country to which they were to transfer their name.

So much for the ethnic movements on the western side of the islands. Those on the eastern side about the same period were also of importance ; but that importance has, I think, been exaggerated. I mean the gradual or successive introductions of Germanic blood into the coast -lands,

THE RACES OF BRITAIN. 35

extending between Portsmouth and the Wash, which Kemble, Palgrave, and several other modern writers on this portion of our history, in- cluding all those in whom scepticism is so strong that the}- prefer conjecture to tradition, regard as having amounted to a thorough Germanisation of the whole region, so as to entitle it to be called " the Saxon shore " by as full a right as it had acquired two hundred years later. I prefer to range myself in this matter on the weaker side, if that can be called the weaker side, which, though few in number, includes the powerful authority of Dr. Guest.

The designation of Litus Saxonicum appears to me no more to imply that the Saxons had occupied the country than that of the " Welsh Marches " implied that the Welsh occupied West Worcester- shire in the twelfth century. There was a corresponding Litus Saxoni- cum in Gaul, in parts of which, and especially about Bayeux and in Artois, Saxons may have been actually settled, nay, did settle, either then or at a later period ; but the name in that instance was extended to the whole coast of Brittany, and we have no reason whatever, historical, linguistic, or physico- anthropological, for supposing that the Saxons ever settled except on very small portions of that coast. It is extremely improbable, considering the troublesome piratical habits of the Saxons, that the Romans would have chosen to constitute a complete fringe of them ::: on a coast abounding with inlets and harbours admirably adapted for carrying on their favourite occupation. But if it is meant that not only the coast, but the inland parts bounded on the west by a line drawn from Portsmouth to the Wash, had been re-peopled with Saxons by the Romans, these further difficulties arise :

Why was the Comes Litoris Saxonici not rather called Comes Pro- vinciae Saxonicae, or by some equivalent title ? When and whence was so immense a multitude of Saxon captives obtained, from a people whose own territory was not then conterminous with that of Rome, and who appear to have confined themselves to maritime expeditions ? How is it that in the Notitia Imperii we find no evidence of the impo- sition of new local names in the interior by the colonists, which, if the settlement had been of the character and magnitude claimed for it, would almost certainly have taken place, considering the passion for bestowing new names which their subsequent history shows the Saxons to have had.

The argument, then, based on the words " Litus Saxonicum " may perhaps be dismissed ; but there are other grounds for supposing the existence in Britain of a considerable Teutonic population introduced by the Romans. In the first place, as Wright remarked, the longer the empire endured the more dependent it became on the swords of the Germans ; and so that race may gradually have come to predominate in

* According to the Notitia, the second legion lay at Richborough, and auxiliaries in the other garrisons along the Saxon shore, Dalmatians, Sarmatians, Belgic Gauls Tungrians, but no saxons. — Guest, Early English Settlements in South Britain

-i *

36 THE RACES OF BRITAIN.

the military colonies, in several of which Tungrians and Batavians were stationed from the first. Moreover, we have precise statements as to the introduction of other bodies of Germans on several occasions. Thus Marcus Aurelius brought over a multitude of Marcoman captives, and Probus did the same with such Vandals and Burgundians as he had taken, as well as a number of Alemanni. Crocus, an Alemannic king or chief, was one of Constantine's supporters when he was proclaimed emperor at York, and had doubtless a following of his own nation ; and the Bucinobantes, a small Alemannic tribe from near Mayence, were transported into Britain by Valentinian, and are supposed to have been located at Buckenham in Norfolk, under their chief Fraomar.

These are all the facts of the kind mentioned by Kemble, or, I believe, by any other authority, Saxon, British, or modern ; and I venture to express the opinion that they are, though of course not unim- portant, scarcely sufficient to bear the weight of inference erected upon them. They do not strike me as of the magnitude required to account for a great ethnic change in an extensive country, which seems at the time to have been pretty thickly inhabited ; for they all occurred long after the supposed destruction of the Iceni, and it is not likely that the best corn district in Britain should have been allowed to lie waste during all that time. The tribes mentioned were of the High German division, and traces of High -Germanism in England are indistinct." Some might say, moreover, that the somewhat elaborate Teutonic social and agrarian system, which Kemble and Palgrave describe as prevailing in Saxon England, and as having been imported from the continental Saxonland, could hardly have grown up, or been reproduced, among a number of colonists scattered in a strange country among alien neigh- bours, and subjected either to Roman military discipline or to Roman civil government.

This position, however, is completely turned by those who, like Seebohm and Coote, deny that the English land-laws and social organi- sation (apart from that of a mere military caste) were German rather than Roman, and who, accordingly, hold that the majority of the popu- lation of Eastern England descend from ancestors who were here before the Anglo-Saxon conquest. These opinions are supported with great ability and learning, and must be further discussed bye-and-bye.

Meanwhile, we may suppose that towards the close of the Roman period — let us say in the year 400, between the usurpations of Maximus and of Constantine — the Romanised population of the towns, under the influence of bad government, dissensions at home and invasions from abroad, were beginning to decline in numbers and importance; while the British people, who, still retaining their (as I believe Celtic) lan- guage and partially their institutions, occupied most of the intervening country, probably maintained their numbers and increased in relative importance ; for in the decline of civilisation it is the most cultivated * See, however, Seebohm, especially as to the word Gebur (= villan), p. 394.

THE RACES OF BRITAIN. 37

portion of the community which suffers most and relatively dwindles. There may have been, in some parts of the East of England, limited Teutonic settlements still retaining the memory of their ethnic character, and possibly their language ; if so, they may have been ready, as the Welsh say the Coranied were, to side with Saxon invaders ; but evidence of the fact is awanting. In the western parts the Roman power had probably' waned still more than in the east and south : the Celtic tribes were doing pretty much what was right in their own eyes ; and the extrusion or subjection of the Gaelic inhabitants of Wales by the Cymric tribes of the north had probably already begun. With migra- tions under Maximus, or subsequently, from Britain to Armorica, we need not much concern ourselves : they would be rather an overflowing than an extrusion of the native race ; but the uniform statement, found not only in the suspected and much-abused Gildas, but in every native reference to the period, that the Romans ultimately departed from Britain, cannot be neglected altogether. We need not and cannot suppose that there was an universal emigration of the citizens. London and Ande- rida, for example, were well inhabited in the next generation ; and so too, it would seem, were some towns in Kent and Essex. But the wealthy planters, no longer safe under the government of their mother- country, would in many cases return thither, leaving their splendid villas and estates and the British serfs who tilled their fields ; and with the military force would go the merchants and skilled artizans who ministered to it, and to whom it assured peace and employment ; and thus in a short time the towns would probably be drained of the best portion of their population, including almost all those who were more Romans than Britons. A single generation had not passed away when St. Germanus, in his attempts to evangelise the island, found himself dealing with chiefs whose very names a Roman would probably be unable to pronounce ; though a few families, like that of Ambrosius, might for a generation or two continue to pride themselves on their Roman descent and cultivation.

CHAPTER V.

Gbe Bnolo^Sayon Conquest anfc periofc).

WHO were the Saxons, Angles and Jutes? When and whence did they come ? and to what extent did they modify or displace the previous population of the Eastern parts of Britain ?

The last of these questions includes by far the most interesting and important part of the inquiry we are engaged upon, and it is the one respecting which there has been the widest diversity of opinion ; but we must give a little attention, in the first place, to the other two.

A great deal has been written on the relations of the words Angle and Saxon, and of the peoples designated thereby. Latham, in particular, has discussed the matter, according to his wont, very learnedly ; though it can hardly be said that he has made it much clearer. I am disposed to think, indeed, that he has made it in some points a little more obscure, owing to his too great solicitude about the connexion of particular tribes with particular dialects. The great changes that have taken place in the limits of the several languages and dialects of North Germany since the period when we know, on good historical authority, that tribal migrations had almost ceased in the lands between the Rhine and the Eyder and Trave ; the manner and extent to which the Old-Saxon * and the Danish have yielded to the Platt-Deutsch, the Platt-Deutsch to High Dutch, and the Frisian to all its neighbours, while all traces of the Wendish have silently disappeared, should make us very cautious in arguing on the relations of our Teutonic dialects to those of the mainland, the more so as we have not in general the means of making the comparison between contemporary stages of these dialects. The tribal names we have to do with — those of the Jutes, Saxons, Angles and Frisians — were certainly not applied, by the only ancient authorities we have, with anything like the exactness we expect from them. It has been doubted — hypercritically, I think — whether "Saxon"

* It is quite possible, if, as Mr. Howorth supposes, the Saxons were a tribe of conquerors from beyond the Elbe, while the descendants of the Cherusci, Chamavi, &c, were their lcets or churls, that the language of these latter was always Platt-Deutsch, rather than Old-Saxon.

THE RACES OF BRITAIN. 39

was even a native name ; at all events, it has for upwards of a thousand years been bestowed by the Celts upon populations the greater portion of which do not appear ever to have acknowledged it. The limits of confederacies like those of the Franks, Saxons and Frisians (to which we may perhaps add the Angles likewise, who seem to have sometimes included the Warini) varied from time to time, and by no means always coincided with the limits of the dialects. Hengist (if Hengist ever existed, as I believe he did) was probably a Frisian. The Frisians about Tonning have a tradition that he sailed from that port ; but Bede supposed him to have been a Jute, and the Welshmen, who " fled from him like fire," called him a Saxon ; while the Chronicle, in recording the fact, speaks of him and his warriors as Engle ; so that by one and another authority he is made to belong in turn to each of the four tribes in question. I do not wish to deny the existence of differences, physical as well as dialectic, between the Frisians and the Saxons, understanding by the latter name the modern inhabitants of Northern Westphalia and Hanover, &c, Holstein and part of Sleswick. But Zeuss looked upon those of the latter kind as mere developments. He says: "Die Angel-sachsen und Friesen noch Sprachdenkmaler haben, die mehren Jahnhunderte hinaufreichen, und unter sich in naher Verwandtschaft, ferner dem Oberdeutschen oder Gothischen stehen. Angelsachsisch und Altfriesisch sind als spatere Fortbildungen aus gemeinschaftlichen Grunde, dem Ingcevischen Sprachzweige, zu betrachten." Even now the likeness of some of the Frisian dialects to " Saxon " English and to Lowland Scotch is extremely close, though the Frisian naturally retains more archaic forms. It may be worth while to note that the name seax for a knife, said to have been peculiar to the Saxons, is found in the Saterlandish as sox, by the side of knif.

The Frisian physical type is one of great interest to the anthro- pologist, partly on historical grounds, partly by reason of the remarkable and almost peculiar media among which they live. They are an extremely fair and very comely people. Professor Johannes von Muller once made some remarks to me on this point, noting particularly the beauty of the women, which, he said, had struck him immediately on crossing the frontier from a Saxon district. My own observations corroborated those of von Muller. I found the Frisians, from the Zuyder Zee, through Groningen (a Saxonised district), to beyond the Ems, a taller, longer-faced, more universally blond and light-eyed folk than the Saxons, the latter being very often hazel-eyed, even when their

Official (children). Beddoe (adults).

Brown Eyes, per cent., with

*■** { Kg ££::::::::: £ &— «*« >** &-!■■.{ g^^::.1^::: %

Tpp, ( Light hair 944 r ( Chestnut, &c 8-5

Leer (Dark hair 6-84 Leer ( Dark hair 58

Munster District -.{ggfJS:." 18 funster District {SSf&^LZZL t*

4-0 THE RACES OF BRITAIN.

hair is light. :: Dr. D. Lubach, who, as a Hollander, has had greater opportunities than I, distinguishes a Frisian from a Hollandish or Low Dutch type, assigning to the former a taller and more slender frame, a longish-oval flat skull, with prominent occiput ; a long-oval face, with flat cheek-bones ; a long nose, straight or aqiline, the point drooping below the wings ; a high under-jaw, and a well-developed chin ; the skin very fair ; the hair of all colours, but seldom very dark. To his Low- Dutch type he gives a shorter, more thick-set frame, with shorter neck and broader shoulders ; a rounder, broader, less flattened head, little or no projection of occiput ; a rounder face, more prominent cheek-bones ; a nose short, low-bridged ; the chin various, sometimes receding and pointed ; the skin and hair more often dark than in the Frisian type. In the former I recognise the Frisian of D. Macintosh, " not less English," says he, " than the Saxon," and very common in the north of Kent ; the latter is comparatively rare in England, but spreads upwards along the Rhine, and furnished studies of boors to Ostade and Teniers.

The cranial form of the Frisians has been studied with great care and labour by Virchow. Its most uniform and remarkable character is its lowness, its small vertical elevation. It seems, as Virchow himself says, that the brain develops somewhat in breadth to make up for this ; for though, in passing through the country, one gets the impression that the people have long and narrow heads, and though the facial features and the general form of the head itself are of the dolicho- cephalic cast, yet the breadth index, in both old and recent skulls, seems to average as high as 77 or 78, and is often much higher. Withal, the length and narrowness of the nasal opening is remarkable and pretty constant, according with Lubach's description of the nose in the present population, but differing from that of the Merovingian Franks.

As for their relations with the Saxons proper, and the degree of probability of their having united in the conquest or settlement of England, it should be noted that some three or four centuries later than the supposed date of Hengist's invasion, a sort of tribal antipathy seems to have existed between the Frisians and the Saxons, which continued in being, more or less, for centuries after Witikind's Saxons had cut off a Frisian army levied by Charlemagne for foreign service. But there is no evidence that I am aware of of any such mutual hostility at an earlier period than the eighth century as would have prevented some com- munity of action in the invasion of England. Combination among Germanic tribes for similar purposes was notoriously common about the time of the wandering of nations ; and the confederates were often separated far more widely in language and customs than the Saxons and Frisians can possibly have been. Thus, the Cimbri,

* See my tables for Holland and Germany. This difference is confirmed by the more recent official colour-census of Germany.

THE RACES OF BRITAIN. 41

Teutones, Tigurini, whatever they were, must have differed inter se : thus, Ariovistus led several tribes to the field against Caesar ; and, later, the Alans accompanied the Vandals, and the Huns furnished cavalry to the Goths.

Who, then, were the true Saxons, excluding any Frisians who may sometimes have passed by that name ?

There are two or three possible hypotheses respecting them. The name may not have been native, but may have been one fixed on them by others ; in which case it is easier to believe that the Frisians were often included under it. Secondly, it may have been applied to the Cheruscans, Angrivarians, and neighbouring tribes, at a later period than that of Varus and Arminius. This is not very likely, as the Saxons were already known to Ptolemy, who placed them much further to the north-east than the Cherusci. Thirdly, they may have been a martial and aggressive tribe (as we know they had the reputation of being), and may have spread from the neighbourhood of the Elbe over the Weser country, subduing the prior occupants, and becoming the dominant class. This latter opinion, which is that of Mr. Howorth, appears, on the whole, most probable. It accounts best for the disappearance, or perhaps rather the non-appearance, of the Anglo-Saxon speech in the region which the Old English called " Old Saxony," which they erroneously looked upon as their old home because their kindred had come to occupy it since their separation.

We are told that the Thuringians, after their subjugation by the Franks (aided, apparently, by the Saxons), were reduced in status, their nobles becoming simple freemen, their freemen descending to the rank of lazzi or serfs. Thus it may have been with the tribes of Germanic race who were conquered by the Saxons or by the Jutes. I am more disposed to look on the Kentish loets as freemen of the lower rank, who accompanied their lords to England, than as being en masse prior occupants, Romano-British or Belgic.

The true Angles seem to have been the most northern tribe of the Saxon confederation, and to have dwelt beyond the Ditmarshian Frisians, to whom or to the Frisians of Western Sleswick the}' may have been indebted for the means of passage to Britain, though some of them may have extended as far as the Elbe, and thus possessed an independent way of access to the North Sea. The Saxons who were not Angles may have had similar relations with the Southern Frisians ; but, from the description we have of their daring seamanship, it is probable that in the fourth and fifth centuries they had direct access to the sea, possibly by the Elbe if not the Weser. Mr. Kemble found near Stade (a very Frisian region, anciently Chaukian), and also far up the Weser, certain mortuary urns, rare or unknown m other parts of Germany, but known to occur in Suffolk, Warwickshire, Derbyshire, the Isle of Wight, and other parts of England. They may, perhaps, have belonged to a particular tribe of Saxons, for nobody, I believe, locates

42 THE RACES OF BRITAIN.

any Angles on the Weser ; but it is noteworthy, as indicating the admixture of the conquering tribes in England, and the community of enterprise just now insisted on, that the districts mentioned are com- monly called Anglian or Jutish.

Who the Jutes were is perhaps the most difficult part of the question. Latham and others prefer to make them Goths from a Visigothic State in Normandy. I believe them to have been a Gothic tribe, a part of the one which gave name to Jutland, but not Danes ; the true Danes were then farther east. That some Visigoths from Gaul took part in the conquest I would readily admit, as I would admit also the concurrence of Franks from the Rhine, and of Saxons from the new settlements of that people in what was afterwards Normandy. But the West Jutland dialect, like all the others that have been mentioned, has its points of close resemblance to current English, and I see little reason to discredit the old view respecting the Jutes. That they were to some extent a distinct tribe is supported by the positive assertions of Bede, by the statements of the Saxon Chronicle, and by the occurrence of similar and peculiar forms of ornament (the Kentish fibulas) in the settlements ascribed to them. On the other hand, the similarity of the mark, or patronymic place-names, and the other place-names in Kent and in other parts of England, forbids us to imagine an exclusive Jutish nationality."

I have already spoken of the physical type of the ancient and modern Frisians : that of the other old inhabitants of the region, whence the invaders are supposed to have come, remains to be studied. The pre- historic material is, unfortunately, scanty. That dolichocephali existed between the Rhine and the Trave and Oder, as they did to the west and to the east of those limits, may be taken for certain : the discoveries at Minsleben, for example, are sufficient to prove this ; but whether brachycephali were contemporary, though probable, is not proven. They were present in Belgium, we know, at a very early period. Dupont's discoveries, and, above all, Arnold's Sclaineux skulls of the neolithic age, with their breadth-index of 81 to 88, furnish us, probably, with a reason why the modern Walloons have rather broad heads, as well as some modern Zealanders. We have already noted the occasional occur- rence of very broad heads among the Frisians (Virchow) ;*but whether these should be taken to be other than the frontier instances of a variable type, or the results of the influence of a peculiar medium, the alluvial marsh-lands of the North Sea,t on the blond long-headed type, may be doubtful.

Further to the north, in the direction of the possible home of the Jutes, the skulls that have come down to us from the bronze and early

* A.D. 586, Tytila, the son of Uffa, ruled in East Anglia. His name is very like that of the Ostrogothic king Totila. Were the Uffings Jutes ?

t Is it inconceivable that some defect of phosphatic salts in the soil, and thence in the food, might have some influence in this direction ?

THE RACES OF BRITAIN. 43

iron ages are long and very narrow ; ;: and probably this would be the prevailing type among invaders from that quarter ; though doubtless the short-headed type, with high vault and prominent brows, must also have been in existence there, subject or incorporated.

To the south of Jutland and Frisia, in the remaining parts of Holland and of North-Western Germany, though the countenance and general aspect of the people does not vary much in different districts, the breadth-index does, apparently, vary more than in some other countries. Sasse and Lubach, operating on considerable numbers of skulls, up to 20 from one locality, have found the average index vary from 75 to 80 or more, not to speak of the remarkably round heads which occur about the mouth of the Scheldt. + The breadth does not increase as one proceeds from north to south, within the limits just mentioned ; the Flemings, at least, have an index of about 76. (Houze, Virchow.) For the southern part of the Old Saxon country — Westphalia, Brunswick, &c. — there is, I believe, very little information. In 13 living men of Northern Hanover and East Friesland, I found a mean index of 79/2, which in the skull would be about 77.

So far, the modern inhabitants of these countries would appear to differ little in the most important article of head-form from those of the East of England, though they may have in some districts a tendency to greater breadth. The race elements cannot have altered much since the era of the Saxon conquest : the only new ones worth mentioning that have been introduced having been, firstly, the Slavonic captives, reduced to slavery and brought westwards during the Germanisation of the country beyond the Elbe ; and, secondly, the refugees from the south during the religious wars. These elements, if they had any effect at all, would tend to enlarge the prevailing indices.

YVe have, however, fortunately, a valuable accession to our know- ledge of the physical type of North-West Germany in the paper of Dr. Gildemeister, in the Archiv fur Anthropologic for 1878. It embraces full details of the measurements of 103 skulls, from old burying-grounds at Bremen, some dating from the beginning of the Christian period, and the whole yielding, says Gildemeister, " ein Bild der Bevolkerung Bremen's etwa zwischen dem 9 und 14 Jahrhundert." One would expect to find a (so-called) Saxon population, with some Frisian blood, but very little other admixture.

Gildemeister divides this remarkable collection under three types, assigning 72 to the grave-row (Reihen-Graber) series ; 26 to what he supposes to be a Frisian type, but which, unfortunately I think, he styles Batavian ; 5 to a brachycephalic one, unnamed. I do not feel sure that the division between the first and second type is fully warranted, for

* Virchow, Archiv. f. A. 1873.

t At Saaftingen, below Antwerp, the index of 1 1 skulls from a submerged church- yard was found to be 86-4, while the same crania belonging to Antwerp yielded one of 77. — M. Kemna, in Bulletins of the Paris Society for 1877.

44

THE RACES OF BRITAIN.

reasons which will be more easily comprehended after inspection of the accompanying diagram (opposite) and table, which I have constructed from Dr. Gildemeister's data :

	46 26 13 5 63 4° 103	S . 11	B O	2-? c a O 0) £a	"5-S	Si CD .E «	a 0 a 0 2 3 0 b	"3> °3 .c u 0! W	Occipital Length. Longitudinal Arc.	a~	M a •a op X	0 -3 a 0 0 z
	190-2	180-5	972	129-3	1397	532-6	118-9	987 377	736	71-6	466
Females	1827	170-8	937	...	137-8		II5'4	... ...	75-2	7i'4	47'9
	192	183-2	98-2	131-4	l5l-4	551-5	116-3	987 379	79-7	661	47'3
	i8o'g	170	97' 1		143-1		m-3			79.2	68-i	514
Brachvcephal (Males?)	1 75 '4	170	99"4	134-8	I53'4	524-2	121	93	367	869	75'4	45
Whole Series, Males	1 90' 8	181-3	97-6	130	143	5366	118-9	98-7	378	75-8	70-8 46-6
	i8r8	170-5								766	70-5 49
Both sexes										76-i	7°7| —

When the breadth-indices have been arranged diagrammatically, after the method which Kollmann and Galton especially have utilised, it will be seen that, though the series is somewhat irregular, it does not depart very greatly from the line of a curve whose maximum point would be about 77; not very much higher than the average, which, it will be seen, would be about 76-1. But when the series is dissected into three, though the small Batavian division yields a tolerably uniform curve, the Graverow one, which, after culminating at 75, should gradually descend to zero at 85, comes to an abrupt conclusion at 79-80.

If we construct a diagram of the height-indices, the result is much the same ; there is a much nearer approach to a regular and satisfactory curve if we throw together the three divisions of Dr. Gildemeister. On the other hand, it must be confessed that variations so great as from 66 to 89 in the index of breadth, and from 59 to 82 in that of height, are very unusual in an unmixed race ; in which, indeed, the highest possible authority (Broca) says they should not exceed something between 10 and 15. The differences between the Graverow and Batavian divisions which appear in the measurements are not very great : the latter are more exuberant laterally, in the temporo- parietal region, distinctly lower, and somewhat more capacious ; but in occipital projection and breadth of base there is hardly any difference. The 5 brachy- cephalic males are so different from the " Batavians," in almost every point except the absolutely large breadth, that if we concede a several origin to the latter, we certainly must do so to the former. In that case, they may probably be referred to the Disentis type of Switzerland and Swabia.

These two subdivisions excluded, the whole remainder, nearly three- fourths of the entire series, may be taken as the mediaeval representa- tives of the Saxons of Witikind, if not of Hengist. They pretty closely

NUMBER OF SKULLS. 56 7 8 9 10

U

o

o 33 >

< m

30

O

<

■0 m

r o w

l-H

H

(75

(7)

a r r

w

46 THE RACES OF BRITAIN.

resemble the remains we possess of the other Germanic conquerors of the fourth and fifth century, the Merovingian Franks, * the Burgundians, the Alemanni at an early date when little mixed, and even the Goths, if the cemetery at Almunecar be really Gothic. The resemblance is not one merely of breadth, but extends to the general contour, both facial and cranial. Very constant are the projection of the occiput, the post- parietal flattening, the somewhat roof-like norma occipitalis, the round- ing of the frontal region as viewed from above, the prominent chin, the nearly vertical cheek-bones. Small differences there are, which may have been tribal, though certainly not constant. Thus, the nasal index is said to be larger in the Franks (as also in many of the modern Hollanders, who may represent the Salian Franks) than in the Graverow men of Swabia, the oldest Alemans. In Gildemeister's Saxons it will be seen to be small (46-6 in the males) : what it was in the Anglo-Saxon I regret to be unable to affirm with any confidence ; the point could be settled only by very careful and leisurely measurement. There are differences also in the degree of development of the superciliary ridges, which may have been more tribal than individual ; and here Gilde- meister's Saxons, and the Saxon conquerors of England, were perhaps, as a rule, less rugged in feature than the Franco-Thuringians. On the whole, I believe the Saxons, Franks, Burgundians,! and Alemanni to have been branches, but lately separated, from one great blond and long-headed stock, which may have been the Hermionian of Tacitus. Differences of dialect, known to have existed at a later period, but fairly conjectured, from the personal names, to have been already in being, may have been developments partly due to the conquest of, or inter- course with, people of other tongues. In this department, philology may be trusted to follow in the steps of anthropology, so soon as these shall have been firmly planted. Anthropologists had long been crying out for the remains of an Iberian or pre-Celtic language in the British Isles before their philological brethren woke up to the consciousness of their existence. Mongolian or Ugrian types had been recognised, though less distinctly; and now Ugrian grammatical forms are being dimly discerned in the Welsh and Irish. \

On the whole, the following theory of the origin of the "Anglo- Saxons" may be put forth as probably not far from the truth :

The first invasion was made in Kent by certain of the northern Frisians and their neighbours the Jutes. Its success encouraged the more adventurous spirits in all the country near the coast, from the Rhine to beyond the Eyder, to follow their example. As a rule, the more northern of the invaders attacked the more northern parts of the

* Burgundians of Savoy, lat. ind. 749 (Hovelacque) ; Merovingians, 763 (Broca). The figures given by Ecker, His, and Von Holder, for the Alemanni, are much the same.

t De' Mortillet says the contents of Franks and Burgundian tombs are almost identical ; the same brooches, beads, belt-plates, pottery, &c. The Saxon paraphernalia are nearly the same J Elton, p. 1C7.

GILDEMEISTER S GRAVEROW TYPE.

sTTs

\>~6Ci-od

GILDEME I STER'S GRAVEROW TYPE, BRE MEN.

GILDEM EISTER S BATAVIAN TYPE.

ARROWSMITN BRISTOL

Ill
Q.
>
h
£
o
tr	a
UJ	UJ
?	0)
tr
o	u
	s
lL	Ul
o	a
	_i
^	5
h
UJ
E
<
>

THE RACES OF BRITAIN. 47

island, landing about the Yare, the Wash, the H umber, or even the Forth ; and these caused a preponderance of the Anglian dialect, which leaned somewhat towards the Danish, in those regions ; while the southern pirates, from the Rhine, the Yssel, the Ems, the Weser, and, above all, the Elbe, occupied the south-eastern parts of the island. Certain Visigoths from Normandy may have joined with the Jutes, who, by hypothesis, were a remnant of their kindred who had never hitherto left the neighbourhood of the Baltic, in supporting the great invasion by Cerdic ; and they acquired lands in the Isle of Wight and the district of Meon ; but as the southern or true Saxons much predominated in that quarter, they ultimately adopted the dialect of their neighbours. :|: Spite of any possible shades of difference in language, the evidence of a common national feeling among them, a consciousness of racial identity, was very strong. In Beowulf even the Danes are spoken of as if of the same race. The national traditions and national heroes were common ; the kings drew, or believed they drew, their origin from the same sacred family; their religion was the same ; their laws differed little. Not that in these half-military, half-piratical, colonisations this kind of congruity among the invaders was likely to have been much considered : the principle of selection, if selection there was any, was probably that of the Jomsburg vikings, who cared little for nationality in comparison with prowess, f

The date of arrival of the Saxons (or Jutes) in England has been much debated. I trust I have shown some reasons for my belief that only piratical descents, not leading to permanent settlements, had taken place before the abandonment of Britain by the Romans. How soon after that event the Saxon colonisation began is a matter of less importance. Haigh brings forward a good deal of evidence to show that it was in 428 ; but it may have been as late as 449, the commonly received date.

In considering the manner and extent of the earlier conquests of the Saxons, as I shall continue to style them in order to avoid any ambiguity that might arise from calling them English, one finds oneself compelled to make an estimate of the credibility of the only author who could have written from personal knowledge or direct information, namely, Gildas. Gildas has been fiercely assailed of late years, chiefly by those writers who hold strong views as to the predominance of the British element in the modern English people. No doubt his work is liable to the strongest objections : meagre in fact, copious in words, stuffed with sounding

* There is an argument against the importance of any Visigothic constituent of the invasion, which I do not remember to have seen noticed. The Goths in France were already Christians of the Arian sect.

t Jarl Hakon, of Norway, after his battle with the Jomsburgers, was presiding at the beheading of his prisoners. His son induced him to spare eighteen of them, among whom was found a Welshman, named Bjorn. The fact is a warning not to trust too much to the linguistic character of proper names. (Note by Lord Strangford.— Welsh- man is as likely as not to mean Italian or Frenchman, without any further specification. Bjorn may have been an Orsini from Rome.)-

48 THE RACES OF BRITAIN.

epithets, it is evidently the production of a monk who, himself more than half a barbarian, outdid, through affectation, the faults of civilised writers. Some portion of what he relates is, no doubt, incorrect and exaggerated, but there seems no reason to suspect him of wilful falsifica- tion. Unfortunately, his aim was not so much to give us the historical facts of the conquest, as to improve the whole affair for the benefit of his countrymen. The book must be taken for what it may be worth, which would not be very much, indeed, if we had anything better to substitute. It contains one famous passage/' which, having been quoted by both Bede and Nennius, has greatly aided in bringing about the popular belief that in most parts of England the Britons were either exterminated or utterly enslaved. To the consideration of this I shall return after a while.

Our other principal authorities for the Saxon Conquest are Bede, Nennius, the Amides Cambrics and Bnct y Tywysogion, which, however, are extremely meagre on the points which concern us ; the genuine poems of Aneurin and Llywerch Hen, and, above all, the Saxon Chronicle. To these may be added the works of numerous chroniclers who founded their works mainly upon the last-mentioned storehouse of facts, including Ethelwerd, Henry of Huntingdon, and many others. In some cases these later chroniclers seem to have had access to sources of information lost to us ; but in other of their amplifications it is pretty clear that they sacrificed truth to what they thought literary effect.

Putting together what we can gather from all these writers, we obtain a pretty fair account of the gradual conquest of Wessex, and some idea of that of Northumbria, and even of Sussex and Kent ; but as to that of the rest of Eastern Britain we are left in the dark in a most remarkable manner.

The great gaps in the story might be partly filled up if we could, with Air. Haigh, trust in other authorities which are generally regarded as spurious or valueless. I confess that the work of this laborious author has educed from what at first sight appears a mere chaos and farrago of invention a certain number of not improbable conjectures. I would not be too ready to deny that there may have been a period of savage warfare preceding that of settlement, during which the whole east of Britain may have been ravaged, and in which the Frankish king Childeric may have played a considerable part. But on the whole the ordinarily received view, shorn of certain features which are obviously conjectural and untrustworthy, such as the deeds of Port and Wihtgar, seems more probable.

Kent, then, appears to have been conquered not at a stroke, but gradually, after several battles in which the invaders were not uniformly

* Itaque nonnulli miserarum reliquiarum in montibus deprehensi ilceroatim jugula- bantur ; alii fame confecti accedentes, manus hostibus dabant in cevum servituri, si tamen non continuo trucidarentur, quod altissimas gratia? stabat in loco; alii trans- marinas petebant regiones, cum uludatu magno, &c.

THE RACES OF BRITAIN. 49

successful. The conquest of Sussex was also gradual ; the natives were in one city at least utterly extirpated, but they seem to have retained their hold on the great forest of Andred for a further indefinite period. Meanwhile settlements were being made all along the eastern coast, but no important states were constituted. Towards the end of the fifth century Cerdic (?) and Cynric began the reduction of Hampshire and Berkshire, which were the nucleus of the important kingdom of Wessex, and they even occupied some portions of Oxfordshire ; while Surrey, so far as we can see, may have been still British. Meanwhile Essex was being settled. From the evidence of mixed Saxon and Roman occupa- tion, such as coins, &c, in some towns of Essex, it may be that this conquest was not so much a war of extermination as that of eastern Wessex seems to have been. Towards the middle of the sixth century Ida began in earnest the reduction of Northumbria, on the coast of which Anglian and Frisian settlements had been going on for some time ; he overthrew the Britons in the great battle of Cattraeth, and appears to have displaced and compelled southwards some of the tribes that opposed him. In the latter part of the same century the monarchy of East Anglia is said to have been formed ; but probably the North and the South Folk had previously constituted independent tribes. The vigorous line of princes who ruled in Wessex meanwhile pushed their conquests northward and westward, carrying their arms as far as Bedford, which was still Celtic (571), and acquiring (577) all the country up to Gloucester, including that town. How far beyond this Ceawlin extended his dominions is doubtful ; but there is reason to think that he overran most of the valley of the Mid-Severn, which, however, must have been lost to him, but not necessarily lost to the Saxons, by the defeat of Wodnesburgh in 591. Dr. Guest ingeniously showed the pro- bability that the Damnonian kingdom for some time after these events still included the upper valley of the Bristol Avon and the western borders of Wilts. The history of Wessex is thenceforward that of a succession of contests against the Damnonian Welshmen, and occa- sional extensions of the frontier by further conquests from them. In 658 the Welsh were driven across the Parret, which seems to have been the boundary for some time afterwards, though in Ine's time (722) Taunton was already within the Saxon frontier. The conquest of Dorset is obscure ; that of Devon was probably not effected till the time of Egbert, and cannot then have been altogether complete, though the statement that the Dene fought the Wealas at Gafulford must be taken to imply the existence of a Saxon military colony, at the least, in Devon. .The reported birth of Wilfrid at Crediton, in the seventh century, is worth noting. Possibly small local settlements of Saxons on the banks of the navigable rivers preceded the conquest ; certainly my figures indicate that a blond population is found precisely in those quarters. Still, there was a half-independent British community in Exeter till the time of Athelstan, who "drove them "out," i.t., probably, put an end to

5

50 THE RACES OF BRITAIN.

the corporate or separate legal existence of the Welshmen, both there and elsewhere east of the Tamar. In Cornwall they retained some shadow of independence yet awhile.

So much for the history of the South of England, as it bears on our subject. That of the north is less clear, and that of .the centre very scanty. Theodoric, son of Ida, is said by Nennius to have been besieged by Urien, Morgant, and other Welsh chieftains, in Holy Island, for three days. This may imply that the British frontier in Bernicia had not yet been thrust back so very far from the coast ; other- wise the Anglian king could hardly have been so successfully surprised. Ethelfrid Fleisawr, the ravager, who acceded in 592, seems to have extended greatly the limits both of his kingdom and of Anglian coloni- sation. " Rex fortissimus et gloriae cupidissimus," says Bede in an important and often -quoted passage, "qui plus omnibus Anglorum primatibus, gentem vastavit Britonum. . . Nemo enim in Tribunibus, nemo in Regibus, plures eorum terras, exterminatis vel subjectis indi- genis, aut tributarias genti Anglorum, aut habitabiles fecit." As only about sixty years intervened between the death of Ethelfrid and the birth of Bede, this evidence that from some districts the natives were entirely driven out seems to be very valuable. It is probable that Lancashire, Cumberland, and indeed all the country west of the moun- tains, remained altogether Celtic for some time longer, though doubtless they paid tribute to Ethelfrid. The little British state of Elmet, about Leeds, endured till the time of Edwin, who expelled the king, and probably also the inhabitants ; * for their mere subjection would hardly have been noticed among the achievements of a king who, according to Bede, was suzerain of all the Britons, including the men of Gwynedd, Anglesea, and Man. In Edwin's reign, at least during the earlier part of it, the Mercian kingdom had not yet been constituted: this is indi- cated not only by the direct statement of the chroniclers, but by the whole history of the political relations between Edwin and Cwichelm of Wessex. I shall endeavour, hereafter, to augment, by inferential evidence of other kinds, the probability that great part of Mercia was really colonised from Deira, or by way of the Humber and the Trent.

Under Penda, Mercia seems to have reached as far as the Severn, though even this is uncertain. Under Wulfere, a little later, Merwald, his brother, is spoken of as ruling in West Hecana (Herefordshire), but this can only indicate a sort of peninsular extension of Mercia.

The growth of the Northumbrian states was checked for a time by the gallant efforts of the Welsh prince Cadwallon, who was probably the ruler of the Northern Cumbrians as well as of North Wales, i as the scene of some of his battles is placed at least as far north as Tynedale. After his defeat and death, Oswy and Ecgfrid for a time must have held

* The modern inhabitants of Elmet are a remarkably fair race. (See Tables for Yorkshire.)

t This is asserted by certain Welsh authorities. — Cambrian Register, ii., 527.

THE RACES OF BRITAIN.

51

at least the suzerainty over the whole country from the Mersey to the Forth, and from the Humber to the Mull of Galloway ; and it would appear from the statements of Bede that under the latter monarch some portions of the newly subjugated territory were actually colonised. Cartmel in Furness was British ; but the gift of the land by Ecgfrid to the see of Lindisfarne, " with all the Britons thereupon," implies a very complete subjection of the district to the Northumbrian power. So does the establishment of a Bishop at Whithern in Galloway, and the apparently English name of that place. After Ecgfrid's defeat and death among the Picts many of his conquests were lost ; but Herebert, the anchorite, an Angle, dwelt near Carlisle ; and there was a monastery on the Dacre river, near Penrith, whereof the names of the abbot, Suidbert, and a presbyter, Thuidred, prove the Saxon character. That near Hexham English was the language of the lower orders is indicated by the story of the dumb man cured by John, Bishop of Hexham. But in 731 the Northumbrian kingdom still extended its limits beyond those of the English language ; for Bede, writing in that year, remarks of the Britons, '• quippe qui quamvis ex parte sui sunt juris, nonnulla tamenex parte Anglorum sunt servitio mancipati." Another flow of the tide of Northumbrian power took place about the middle of the eighth century, when King Eadbert conquered Kyle (Mid -Ayrshire) and took Dum- barton. From that time until the Danish conquests the independent British dominion was probably restricted to the most rugged and infertile parts of the ancient Cumbrian kingdom, and of these only the districts north of the Solway. But that the population of parts, at least, of Cumberland, Westmoreland, Furness, &c, remained British in language and feeling is rendered more than probable by the reconstitution of a southern Cumbrian state in the latter half of the ninth century, when the English were too much occupied in resisting the incursions of the Danes to defend the outlying portions of their territory. In the Lothians and Tweeddale the Anglian population, once established, always remained predominant.

The state and position of the Mercian and Welsh frontier must have fluctuated considerably from the time of its formation. We have seen that Ceawlin of Wessex may have been the first conqueror of the middle as well as of the lower portion of the Severn valley. Cynddylan, Prince of Powys, is thought to have been the Condidan slain by him at the great battle of Deorham ; and in his elegy by Llywarch Hen are allu- sions to the destruction by the enemy of Cynddylan's city on the Tren or Tern. So, at least, the poem is generally read. Moreover, near the same time, Gwen, the valiant son of Llywarch, was slain at the ford of Morlas, close to the present frontier, though we are not distinctly told that his enemies were Saxon. It has been suggested by Dr. Guest that in the same campaign in which the battle of Fethanleage (Faddiley in Cheshire ?) was fought Ceawlin destroyed Wroxeter, and overran the whole country as far as Cheshire ; and I am disposed to embrace this

5 *

52 THE RACES OF BRITAIN.

view. But the portion of Hwiccia which had been conquered in this and preceding campaigns seems to have broken off the yoke of Wessex in the year 591, at the battle of Wanborough. As this battle was won by a combined force of Britons and rebellious Saxons, we are left in doubt who retained the supremacy over Hwiccia, which, however, was undoubtedly a Saxon land a century later. Its dialect is a southern one to this day, though it remained a sub -kingdom of Mercia after the seventh century. Though there were, as has been already mentioned, Saxon settlers (Magesaetas) in the district between Severn and Wye (Mayhill, between Ross and Gloucester), it was not till the reign of Offa that the Welsh were deprived of all Salop and of Herefordshire north- east of the Wye. Indeed it is difficult to account for the paucity of hundreds, i.e., of free population, and the entire absence of history during the early part of the Saxon 'period, in Staffordshire and East Cheshire, except by supposing that these districts were still little more than border-land's. In the early part of the ninth century the process of pushing the march westwards was still going on. In 823, according to the Bvut-y-Tywysogion, the Saxons took the kingdom of Powys into their possession. But here, as in Northumbria, the further progress of the Saxon arms and frontier was checked by the Danish invasions ; and for two centuries more the Wye and Offa's Dyke may be said, roughly speaking, to have been the boundary of the nations, though Welshmen probably remained east of those boundaries in considerable numbers, obeying English law, just as they did in West Wessex when the laws of Ine were compiled.

Ethnological changes in other parts of the British Isles than England and Southern Scotland, during the period we have been considering, were probably but inconsiderable. Professor Rhys holds that in Scot- land north of the Forth vacillations occurred in the relative position and power of the Brythonic (Cymric), true Gaelic, and Pictish elements, the last of which he supposes, as we have seen, to have been still more or less Iberian or Ivernian in language in the early part of this period, basing his belief partly on the use of interpreters by the Gael Columba in his communications with some of the Pictish people. There can be little doubt, however, that the Caledonians were a Gaelic people ; and that they, who were classed as Picts by the Romans, continued to exist and flourish in the Central Highlands, where their type as described by Tacitus is still nowise uncommon, though intermixed with another which was probably Iberian.

In the west of Scotland the true Scotic Irish streamed over at inter- vals into Argyle and the neighbouring isles ; and the Irish Cruithne, or Picts, similarly crossed over the narrow channel into Galloway, where they, reinforcing probably a pre-existing tribe of their own kindred, maintained a separate tribal existence for some hundreds of years, being still recognisable at the Battle of the Standard in 11 38. In Ireland itself, meanwhile, no new element was added to the population ; nor

THE RACES OF BRITAIN. 53

does it appear that continued intestine warfare brought about any con- siderable changes in those which already existed : only the Cruithne, doubtless, were gradually melting away into the Gaelic-speaking mass.

We may now proceed to consider the evidence to be derived from proper names of persons, from statements and allusions which occur in the documents of the period, and from the nature of laws and social usages, as to the ethnological character of the subjects of the Heptarchy. The evidence of local names, and the most important of all, that of physical type and characteristics, must be reserved for a later stage.

One is struck with the extreme rarity of names apparently British, not only in the history, but in the documents of the whole period. Kemble has found a few in the earliest charters contained in his Codex Diplomaticus ; but there is nothing discoverable that can be compared with what we find, for example, in Gaul under the Frankish kings, where persons with Latin names very frequently appear, not only as churchmen, but as councillors and leaders.

The name of Coifi, the heathen priest under King Edwin, who was first of the Northumbrians to embrace Christianity, may easily be derived from the Celtic, but hardly from the Teutonic. There is no corresponding fact in Wessex ; but we meet there with one equally noteworthy in another rank, the name of Ceadwalla, the savage con- queror of Wight and Sussex, being clearly Welsh, and its owner being claimed by the Welsh as a countryman and their king, by the name of Cadwallader. There is, however, no sign in the histories of any revolu- tion favourable to the Britons about that time ; and Ine, his cousin, who succeeded, carried on wars of conquest against them, and allowed a smaller weregild for their slaughter than for that of Englishmen. No Celtic name appears in the genealogies assigned to Ceadwalla ; and it can hardly be doubted that, if British at all, he was so by the mother's side only, which would probably be enough to account for his name, and for that of his brother Mul (the mule, half-breed).

There is, however, one portion of Mercia, and by no means the one in which the presence of un-Saxonised Celts might have been looked for from its geographical position, in which two very Celtic-looking names do occur. Ovin is noticed by Bede in the ancient History of Ely, as primus ministrorum of Etheldreda (a.d. 670?), and is thought to have had the administration of the Isle of Ely.:|: His monument is now in the Cathedral there. And when Ethelfled's men, in 916, stormed Derby, one of her captains, of the race of the Lords of Ely (South Gyrwa), and called in the WTelsh annals Gwynan, set fire to the town-gate. If, in connexion with these names, we consider that the Fen country long constituted separate districts, called North and South Gyrwa, but very thinly peopled, having indeed only 600 hydes in each ; when we call to mind the story of St. Guthlac, who, about a.d. 700, was surrounded by night, in his cell near Croyland, by a crowd of enemies, and from their

* Palgrave, II., ccciii.

54 THE RACES OF BRITAIN.

rough and guttural speecli imagined them to be Britons, by whom the country was then much harassed, though they turned out, to the saint's great relief, to be only devils, and not Welshmen ; and when we find in the history of Ramsey" that that neighbourhood was, even in the time of Canute, liable to " the infestation of British thieves," and that in the regulations of the Thegna gild of Cambridge, probably in the reign of Ethelred, provision is made for the slaughter of a Welsh churl, and that at half-price, we shall probably agree that a case is made out in favour of the Britons having remained as Britons in the neighbourhood of the Fens much longer than in most parts of England. I hope to show, further on, some little evidence of a physical character in support of this view.

The use of the words Wealh (a Welshman), Wylisc (Welsh), &c, is not infrequent in the ancient Saxon laws. These words do not, how- ever, occur in the ordinances of the Kentish kings, though the frequent mention of " esnes " and " theows " shows that the servile class was numerous. One would infer that the Welsh language, and with it the distinction of race, had become extinct by the time of King Ethelbert, which is inconsistent with the notion of any considerable portion of the native population having been suffered to remain in Kent by the con- querors. If Mr. Kemble be correct in his reckoning of the Kentish weregilds, the Kentish ceorl was valued at half the worth of an eorl ; whereas in Wessex the ratio of values was as i to 3 at least, or 6 in the case of the full thane ; in Mercia, as 1 to 6 ; and in Northumbria, 1 to 7^. These differences may be connected with the relatively higher value of the simple freeman among the Frisians and Franks! than among the Saxons and Angles, whose original constitution appears to have been more aristocratic. But, at all events, the relatively high weregyld of the Kentish churl makes it very improbable that he was, as a rule, of British descent.

We have some valuable material in the often-quoted laws of Ine of Wessex, who lived at a period when the prevalence of Christianity had in some small degree mitigated the horrors of wars of conquest, and who had himself reduced West Somerset to be part and parcel of Wessex. Under Ine a Welshman might be free, might own or rent land, and might by holding five hydes attain the rank, or rather value, of a sixhyndman. We hear of no other sixhyndman, as Robertson remarks ; and it may be that he was in a class by himself, below all thanes of true Saxon blood, who were twelfhynd. A distinction was still drawn, too, between the English and the Welsh churl. The life of

* Quoted by Palgrave.

t The weregylds in Kent (see Robertson, Scotland under her Early Kings, for the most comprehensive statement of weregylds) seem to agree in proportion with those of the Salian Franks, who probably took some part in the colonisation of that county. Grimm, quoted by Kemble, makes those of the continental Saxons resemble those of our Angles; but those of the Thuringian Angles differed from both. For an indication of the greater political importance of the churls of Kent, see Thorpe, Laws of Athelstan, II.

THE RACES OF BRITAIN. 55

the former was valued at 200 shillings ; but that of the latter, though he held a hyde of land, was but 120 shillings; and if he had none he was rated at 60 shillings, the price of a theow or slave. We may infer that in a considerable portion of the Wessex of that period (about a.d. 700) Welshmen and Englishmen lived intermixed under English law ; and though the former, as the conquered race, were subjected to some dero- gatory legal provisions, they were not altogether deprived of their lands, and remained too numerous and powerful to be treated without some consideration. It by no means follows, however, that this intermixture extended to the eastern portion of Wessex, which had been conquered before the introduction of Christianity, when community of religion had not begun to soften national antipathies. In Mercia the eorl was worth six times as much as the ceorl ; the probability of the latter being generally of British origin is somewhat greater, therefore, than in Eastern Wessex. From Northumberland we have precise information in the fragment entitled "The North People's Law," * which, however, Kemble, without reason assigned, seems to consider a suspicious authority. It sets the value of a churl's life at 200 shillings, while that of the lowest rank of thane is not less than 1,500. This comparative elevation of the upper ranks may have taken place subsequently to the Danish conquest ; for it is not probable, from the nature of that con- quest, that any considerable number of churls were Danish. The soc- man does not appear at all under that name ; he ma)' have been a development of the Danish leysing or of the churl. Welshmen are provided for almost as in Wessex, a landholder of that race being worth 120 shillings, and a simple Welsh freeman 70 shillings ; a churl may rise to be a thane, but it is not stated that a Welshman may do so. These provisions for Welshmen are to be accounted for by the continued existence of the British language as well as the British race, under Danish and Norse as well as Anglian Northumbrians, in Cumbria and Craven. This latter district, from which we have evidence of a singu- larly interesting kind, derived from the contents of caves, that civilised Britons took refuge there from Anglian or other invaders, is generally supposed to be the commot of Carnoban, spoken of in one of the Triads as having remained British in speech. The name Carnavy or Cornaby, said to be applied by the peasantry of Warwickshire to a district in the centre of that county, raises a suspicion whether the commot of Carnoban may not have been in the forest of Arden rather than in the wilderness of Craven, t One is tempted to refer the name to the tribe whom the Romans called Cornavii, and who are located to the east and north-east of the Severn. The population of both these districts retains decided marks of pre- Anglian descent.

There is a passage in the law of the Northumbrian priests (a code also of date subsequent to the Danish conquest) which has been held

* Thorpe, i., 186.

t Carnoban in Deifyr, says the Triad. ' But " in Deifyr " may be a gloss.

56 THE RACES OF BRITAIN.

to imply the continued presence of Welshmen, acknowledged as such, throughout the whole country, a tiling on other grounds very impro- bable. "If a King's Thane," it is said, "make denial, let twelve be named to him ; and let him take twelve of his kinsmen, and twelve Waller- wents ; and if it fail," &c, &c. "If a land-owning man make denial, then let be named to him of his equals as many Wents as to a King's Thane," &c. ; " if a ceorlish man make denial, then let be named to him of his equals as many Wents as to the others," &c, &c. Thorpe supposes these Waller-wents to have been Britons of Cumbria ; but it seems obvious that an inhabitant of the centre of Deira (and York is mentioned in the law) could not have been expected to import twelve Britons from Cumbria, beyond the mountains, to be his jury or compur- gators. Bosworth says: " Wallerwents, peregrini, i.e., Britanni ita dicti a Saxonibus." But as this is the only passage in which the word occurs, it seems to me easier to suppose that it means simply " strangers," "strangers in blood," as distinguished from kinsmen, and has no refer- ence to race or nation, than to believe that there were in the Danish period, in the neighbourhood of York, such numbers of land-holding Welsh-speaking men that a dozen of them could be readily got together for ever}- legal proceeding. There is no other reference in the code to the existence of Welshmen.*

It is ordered in the Judicia Civitatis Lundonicr, respecting a runaway slave, " that the same be done unto him as to a Wylisc thief, or that he be hanged." It is difficult to credit the existence of Welsh brigands near London so late as the reign of Athelstan. Perhaps it would be better to understand the phrase as meaning " a servile thief," " a slave who steals." Abimelech gave to Abraham, according to a Saxon trans- lation, " oxen and sheep, Welshmen and Welshwomen (wealas and wylna)."f So, when in the Laws of Ethelred + it is agreed that " neither they (the Danes) nor we harbour the other's Wealh, nor the other's thief, nor the other's foe," we can hardly translate the word Wealh otherwise than by " slave." It is a fact of fearful significance that the very name of a brave though unfortunate nation should have descended to mean, in their enemies' mouths, a bondsman ; but we have another example of it in the desecration to the same purpose, by the Germans, of the national name with which the Slavonians had gratified their own pride.

Another document- of great importance for our present purpose is the will of King Alfred, in which he bequeaths to his younger son all the lands he has " on Wealcynne butan Triconscire." Triconscire

* There were several landowners with Celtic names (Murdoch, Gilpatrick, &c.) in Yorkshire just before the Norman conquest ; but the names were of Gaelic, not of Cymric type, except perhaps Maban and Artor ; and there can be little doubt that they had been introduced from Scotland through Lothian, where Angles and Scots were then intermixed.

t Bosworth. I Thorpe, i., 289.

THE RACES OF BRITAIN. 57

seems to have been a part of Somerset, which may very well have been Welsh then as in Ine's day ; but in the enumeration of the other lands "on Wealcynne " are included Stureminster (in Dorset) and Ambres- byrig, which has always been supposed to mean Ambresbury = Ames- bury in Eastern Wiltshire. On this some writers have founded the statement that Wiltshire and Dorset still used the Welsh language in Alfred's time. The premises are not sufficient for the conclusion. East Dorset ma)' have been Saxon, though Sturminster was not so as yet. Or these districts may have been still recognised in common parlance as Welsh country, though no longer under Welsh law or Welsh in speech ; just as all Monmouthshire is commonly spoken of, by Somerset and Gloucestershire men, as being in Wales, though it is legally an English county, and though the Cymric language has long ago receded behind the Usk. Finally, if Amesbury (supposing Amesbury to be the Ambres- byrig mentioned, which seems pretty clear) — if, I say, Amesbury was in the Wealcynne, Chippenham, far to the west of it, was not ; for it was not included in the grant, but subsequently given to his youngest daughter. I am inclined to think that Amesbury, or Ambresbyrig, may have really remained Celtic when all the surrounding region had been Saxonised, much as there is some reason for supposing Glastonbury to have done. Ambresbyrig was a sacred place to the Welsh ; it was the seat of one of their three famous choirs and the royal residence* of Caradoc Vreichvras, and they may have clung to and around it long after they had yielded most of Wiltshire to the undisturbed possession of their conquerors. Dr. Guest has succeeded in rendering it probable that the country about Bradford and Malmesbury remained British after Bristol and Devizes and Mere had been subdued. Alfred's gifts, during his lifetime, to Asser, consisted generally of parishes in the western, i.e., British portion of his dominions ; and there was a certain congruity in this, Asser being a Welshman. Why, then, was Ambres- byric (with Banwell in North Somerset) his first gift to Asser ? On my hypothesis nothing could be more natural.!

Another important field for argument is afforded by the various points of resemblance between the Welsh and Saxon, and, again, the Roman and Saxon legal and social systems. The former part of the subject has been worked out by Palgrave and Kemble ; the latter by Wright, Pearson, and Coote, to whom also Seebohm, though taking an independent line, is in some respects a powerful auxiliary. All these take more or less strongly what may be called the negative, or anti- Germanic, side of the question ; the positive or de novo theory, which denies any considerable survival of pre-Saxon usages, has been main-

* Guest.

t The inhabited valleys of Central and Southern Wiltshire, separated as they are by tracts of upland pasturage, have comparatively little intercourse with each other. It will be seen from my tables that the inhabitants of these valleys differ much in physical type, and probably in blood.

58 THE RACES OF BRITAIN.

tained chiefly by Guest and Stubbs, followed by Freeman and Green. I shall not attempt to discuss the whole of this great and difficult subject, nor even to epitomise it, but only to refer briefly to a few important points.

Palgrave and Kemble, while agreeing as to the survival of usages dating from before the fifth century, diverged in their views as to the nature of the population from whom they were derived. Palgrave believed the Belgians to have been the ruling race of Britain before the Romans, and to have been ethnologically the same as the Frisians and Saxons ; while Kemble, relying on that very equivocal term, " the Saxon shore," supposed the eastern part of Britain to have been thoroughly Germanised under the Romans. Both were, therefore, prepared to make much of the resemblances between the British and the Saxon laws. " These resemblances," says Palgrave, " must be sought principally in the tenures of lands, in the territorial organisation of the country, and perhaps in the constitution of the tribunals which were founded on that division. . . . They agreed in their usages respecting crimes and punishments ; they agreed in allowing the homi- cide to redeem his guilt by making compensation to the relations of the slain ; the}'' agreed in the use of trial by ordeal and by compurga- tion. . . . The question whether such analogous customs be of British or Saxon origin is little more than a mere verbal dispute, very difficult to decide, and perfectly useless when decided." Most of the usages aliuded to in the foregoing passage were common to many of the northern nations, and had probably been the common property of at least several branches of the Aryan stock. For example, the general principle of the were-gyld was common to Saxons, to Cymry, to all the Germanic tribes, as well as other peoples. When we come to look into details, we find them varying widely. The elaborate system of were- gylds in the Laws of Howel Dda bears little resemblance to that of any one of the Saxon codes ; and the complicated territorial division, which is said to have prevailed among the Cymry, but which one is tempted to suppose to have been rather the ideal of a legislator than a system actually carried out in its integrity, can hardly be recognised under the Saxons. Palgrave, indeed, treats the British maenawl as the possible origin of the Saxon township; but as 12^ maenawls = 50 trefs con- stituted a commot, and 25 maenawls a cantred (literally a hundred), it is difficult to see how the divisions in the two cases can be made to run parallel. Four trefs constituted a maenawl ; but if we make the town- ship the equivalent of the tref, our difficulties are not lessened ; and the argument which Palgrave derives from the constant appearance of Four Men as deputies from the township, whom he supposes to have been originally the representatives of the Four Trefs of which the Maenawl was composed, will fall to the ground. If, indeed, this system had really been in force throughout Britain at the era of the Roman Conquest, one would hardly look for its persistence throughout the continuance of the

THE RACES OF BRITAIN. 59

Roman dominion. Again, the supposition that the Saxon conquest was a mere substitution of a new military aristocracy for a native one, or for the Roman rulers lately withdrawn, with little interference with the body of the people, or with the laws and usages that governed them — that, in short, it was very like the Norman conquest — can hardly be entertained by those who have any respect for Mr. Kemble's views as to the nature of the Mark and the constitution of a Saxon settlement ; though Coote and Seebohm'may consistently do so.

The election of the tun-gerefa, or reeve, by the villans (allowing that it existed during the Saxon period) is used by Palgrave to increase the probability of the churls being really the representatives or descendants of a subjugated race ; and he brings forward the analogy of the practice in the Levant, where a Greek community, for example, elects its native magistrate, who is its head and representative in all its dealings with the Turkish Government. The fact founded on is correctly stated ; but as Turkish villages and communities equally elect their own head- men, I cannot see that the election of the reeve is any indication of the alien nationality of the churls : it is equally consistent with that and with the other view.

The name and institution of gavelkind furnish a stronger argu- ment. The usage was general in Wales, as applied to the lands of free tenants, though restrained from operating beyond the third generation ; while in England it was of limited and local application ; nor has it ever been suggested that the Welsh derived the custom from the English. It is positively affirmed, however, by Sumner that the Welsh, though they previously used the custom, took the name from the English. On the other hand, the Welsh language affords a clear and unstrained etymology for it ; viz. : gavael, a grasp, and thence a holding ; and cenedyl, kindred or family. The Irish have both the thing, the name, and a probable etymology (gabal-cined) ; and there is choice of one fair Germanic deriva- tion (gafol, tributum, a gift) ; and of one which seems common to German and Celtic, Gabel or Gable, a fork, the angle of a roof; in both of which the essential idea seems to be that of division. If the Saxons took the usage from the Welsh, how came it that it was adopted in one of the counties which had been most Romanised, and which was earliest and most completely Saxonised, viz., Kent," and not in those whose conquest was late, and conducted with less barbarity and slaughter, such as Devon and Somerset, or those where there is reason to think that the British population long retained their language and distinct national character, such as Huntingdonshire ?

If these considerations, however, are not sufficient to rebut the common opinion on the subject, then we have in the existence of gavel- kind among the free tenants of Kent a strong presumption that the churls who followed this custom were mainly of British descent, and

* Pearson would perhaps have said that it was because the eastern part of Kent was acquired by peaceable cession, and only the western by fighting.

6o

THE RACES OF BRITAIN.

a fortiori that the bulk of the population in most other parts of England is so ; for several lines of proof combine to show that the north and east, at least, of Kent, are among our most Teutonic districts.*

The difference between the Salic villa and the Anglo-Saxon tun, the former not being necessarily subject to any lord, while the latter always was so, has also been used as an indication that the Saxon churls were the remains of the conquered Britons ; but, in truth, the continental Franks, as well as the Frisians, were always less aristocratic in their polity than the Saxons. If we may accept the Saxon tradition as of any value, their churls were really the descendants of the subjugated Thuringians, who had first occupied the lands about the Elbe. The same ranks or castes existed among the Saxons on the Continent and in Britain ; and as the churls bore arms, we cannot doubt that many of them accompanied the nobles in the invasion of Britain. It is necessary that we rid ourselves of the idea that our Saxon ancestors were a kind of democratic community with universal suffrage ; if they ever had been such a community, they had ceased to be so before they emerged into history.

From Angeln, indeed, if we believe Bede, almost the entire popula- tion migrated, and it remained a desert until his day. If, then, a portion at least of the churls of insular Anglia were not of continental descent, what had become of the churls of Angeln ? Had they become nobles, i.e., eorls or athelings, in the new settlements ? It is possible, but extremely improbable. The eorls do not seem to have been numerous in Britain. It would seem that they wore the sword, and that it was usually buried with them ; but Mr. Yonge Akerman informs us that in certain Anglo-Saxon villages, whose cemeteries he investigated, the sword-bearers seemed to have been but three or four in number — quite a small percentage. In the time of Edward the Confessor the land- holding thanes of Somerset (and we have none mentioned who did not hold land) were somewhere between 2 and 5 per cent, of the population ; and the burghers, some of whom may have been of that rank, were 3 per cent. Again, if the greater part, or a very large part, of the British native population became under their conquerors free churls, and constituted the mass of the nation, how was it that the words Wealh and Wylisc came to mean slave and servile ? The word churlish never sank to such signification. This single fact is almost sufficient to show that slavery was, during at least the early stages of the conquest, the common lot of those Britons whose lives were spared. That in some western counties the case was different I am prepared to admit, and even to maintain ; there I believe many Britons to have been ulti- mately enrolled among the churls, and the same may have been the case

* Laets are met with in Kent, and in Kent alone eo nomine. They were of three degrees, and each degree had an appropriate weregyld. They were evidently, says Lappenberg, the lazzi or laeti of the Continent. Their existence in Kent furnishes yet another argu- ment against the British origin of the Kentish churls, except perhaps in the Weald.

THE RACES OF BRITAIN. 6l

in many parts of Mercia. Still, even there many were reduced to serf- dom, as appears by the great number of serfs mentioned in the Domesday account of Western and Southern Mercia, a number varying from 10 per cent, in Northamptonshire to 17 in Salop and 24 in Gloucestershire.

At this point it is impossible to pass by without notice the researches and arguments of Seebohm as to the status of the actual cultivators in Anglo-Saxon manors. His work deals with questions which are strictly social, and only indirectly have an ethnological bearing ; still, he shows pretty distinctly his opinion that the churls, or so-called free cultivators, were from the first fixed to the soil and bound down to labour for the lord of the ham or manor ; that this was the case both in England and in Swabia and elsewhere where free Germanic tribes had occupied lands within the old Roman empire ; and that these semi-servile tenants were, in England, for the most part either Germanic captives settled on the land at prior dates by the Romans, or the descendants of the Belgo- British population whom Caesar and Claudius had found already in its occupation.

Elton, on the other hand, holds to the older view, that the bulk of the Saxon settlers had at first a great deal of that freedom of which they long retained at least the name;* but that, with the increase of power of the kings and great nobles and of the church (we may add, with the increase of population and the lessening of that once constant warfare which made fighting men valuable), the descendants of freemen fell under onerous rents and services. And to this view I am disposed to adhere, with some qualification. It is evident, from the examples so carefully given by Seebohm himself, that in the west, at Tidenham in the Forest of Dean for example, the conquerors must have substituted the manorial organisation for that which previously prevailed, and which continued for hundreds of years to prevail on the other side of the Wye. That the Wye was not the national boundary before the Saxon conquest we can hardly doubt ; the fact that it was not the ecclesiastical boundary until then makes this pretty clear. If, then, Tidenham was placed under a new system at that time, while some traces of an older one were allowed to remain in places further east, e.g., Cirencester, may not the Saxons have acted similarly in many or most of the manors of England ? If, again, the services in different manors, at the time when we begin to learn something definite about their nature, varied greatly in detail, may we not allow the possibility of considerable changes, in the direc- tion perhaps of greater onerousness, between the conquest and the periods in evidence ? Again, Cirencester and Tidenham were both con- quered after the invaders had for some generations been familiar with the country and with the various tenures already existing in it.

It is possible that the services on the royal manors, of which Seebohm

* "Even the cottier," we are told in the Rectitudines singularum pcrsonarum, "pays hearthpenny on Holy Thursday, as every freeman ought to do." But not long after- wards, in Domesday, the liber homo is distinguished from the villan and cottier.

62 THE RACES OF BRITAIN.

gives one instance even in the very Saxon Hampshire, may have been heavier than the average of manors held by eorls or thanes. If so, the tenantry on the former may have been in larger proportion wealhs. Thus, after the Norman conquest, as has been pointed out to me by Mr. Bazeley, Saxon customary tenants remained on the king's land in larger proportion than elsewhere.

But in truth the system of noble tribesmen living separately in the family hall or dwellings, and plebeian or servile cultivators dwelling together in a village, was already in operation in free Germany, accord- ing to Seebohm's reasonable interpretation of Tacitus. In fact it was perhaps older and more widespread than the separation between Germans and Celts.

The fate of the Romano -British cities is involved in the same obscurity that more or less clouds the whole subject. Some of them had certainly been pillaged and desolated by the Picts and Scots ; and others were destroyed, and their inhabitants massacred, by the Saxons. Probably Silchester, Wroxeter, and many others, shared the fate of Anderida, where, we are told, there was not one Briton left alive. From others, with the departure of the old civilization and commerce, the population, where not positively enslaved and driven off, would wander away, or gradually perish by famine and disease, as Gildas seems to indicate. Verulam may have been among these : in the tenth century its spacious buildings were a nuisance to the Abbots of St. Albans, being a haunt for bad characters of both sexes ; but the history of the city had been forgotten. In a few places in Essex, and in Ozengal in Kent, according to Mr. Wright, there is evidence that the natives and the invaders dwelt together for a while ; but in these cases, and in others in which the Saxons occupied and inhabited the ancient sites, the Christian religion died out, and the churches were either destroyed, or were no longer recognised as such ; for when Christianity was reintroduced into London, Canterbury, and York, it was necessary to build new churches. In the north of England, we have the evidence of Canon Raine that every Roman station and house bears traces of destruction by fire.

Wright and Coote lay great stress on the strength and population of the towns in the middle and late Saxon periods ; but we have very few facts indeed to support the notion that this importance had continued through the early Saxon period. The only considerable ones I have met with are, firstly, that Cadwallon was besieged by Osric in York, a.d. 643, and that Bede speaks of the transaction as happening in the municipal town ; and secondly, that the burning of Catterick, by Beornred the Mercian, in 769, is recorded as an important occurrence.

The subsequent importance of some cities is of little moment ; for though the Saxons were not originally a town-loving people, they in process of time found the convenience of market towns, and even of fortified places, just as their continental kinsmen did during the ravages of the Avars and the Hungarians; and such towns grew up gradually,

THE RACES OF BRITAIN. 63

in new as well as in old sites.* Among them was new Cambridge, which was built close to the site of the old Camboritum, a Roman city which we know to have been a " waste Chester." Yet Cambridge had acquired great privileges by the tenth century, and in the eleventh it possessed a thane's guild, evidently, in this instance at least, not a Roman survival.

We cannot, I think, neglect to allow great importance to the constant intercourse which subsisted between England, after its conversion to Christianity, on the one hand, and France and Italy on the other. If the Saxons and Angles were barbarians, they were very capable and receptive barbarians. And whereas the Saxon legal codes and muni- cipal regulations, so much of them as is positively known to us, are of date posterior to that conversion, I should be disposed to refer to intro- duction from abroad rather than to inheritance from the Romano-Britons most of those regulations and terms which savour of Latin origin. The one with regard to which I feel most difficulty is the Bridge-and-burgh- rate, to which landed property was liable. Taking this separately, it is vastly easier to suppose it a legacy from the Roman occupation.

The clean sweep which the invaders made of Christianity in the eastern portion of the island is surely not consistent with the views of Mr. Coote. If the Romano-Britons had lapsed into heathenism it would have been their own heathenism, not that of the rude Saxons ; and except the prohibition of drycraft f (Druidic magic ?) by a Saxon law, we have hardly any indication of this in the east. Again, if Christianity had endured in the towns, side by side with the Saxon heathenism of the rural aristocracy, there would surely have been some intercourse with the Christians of the continent ; but we hear of none : no second St. Germanus seems to have crossed over to comfort the brethren and retain them in the faith. If any cultivated Romans, or Romanized Celts, remained to carry on the traditions of the municipalities, it is strange that we have no monuments or epitaphs referable to the succeeding centuries. All these are only inferential arguments indeed, but col- lectively they surely have some force. Mr. Coote finds in the sixhynd- man of Ine's Laws the Romano-British burgess. But we hear of no sixhyndman except the Wealh who has five hydes, i.e., apparently, the Damnonian chieftain who has become subject to the King of YYessex. The English burgess was probably eorl or ceorl, twelfhyndman or twy- hyndman, according to his birth and descent ; and we find the landed nobility owning houses, and probably occasionally dwelling in them and exercising the rights of chief burgesses, in York and in other towns, in the time of Edward the Confessor.

The present seems a favourable opportunity for considering a subject that might have been taken earlier, viz., the testimony of the Triads,

* " Caistor was a city when Norwich was none,

And Norwich was built with Caistor stone." — Norfolk Rhyme. + Pearson, Early and Middle Ages of England, pp. 46, 47.

64 THE RACES OF BRITAIN.

and of alleged Welsh tradition, to the effect that the Lloegrians became Saxon, and that the Coranied treacherously sided with the invaders. Acknowledging my incapacity to test the authenticity of the documents on which these statements are founded, I confess that I regard the poems of Llywarch Hen as genuine, relying as much on the authority of Skene as on the internal evidence. Now Llywarch speaks of " the circularly compact army of Lloegria " as opposed to that of Cyndylan.

I can quite conceive that the Welsh may have, so to speak, served the Romans and the Saxons, successively, heirs to the Lloegrian name, as being possessors of the Lloegrian country. Instances of this kind of misapplication of names are plentiful. Thus Tamerlane, in his journal, speaks of Sultan Bayazid as the Kaiser of Roum ; and Cappadocia, long the eastern fringe of the Roman empire, retains the name of Roum under the Ottoman government. Thus, again (and this is still more closely in point), the old Breton ballads speak of the Franks as Gauls. And the Welsh, in later days, called the Normans Saxons, as they had called the Saxons Lloegres.

It may be, however, that the foes of Llywarch and Cyndylan may have included a Lloegrian contingent. The Britons who fought under Ceolric at Wanborough, in 591, may have been the same who had joined in the storming and sacking of Wroxeter eight years before ; and I would hazard the conjecture that they may have been the inhabitants of Arden,::: that difficult forest region which to this day, as philo-Celtic admirers of Shakspere are fond of telling us, retains indications of a large British element in its population.

But it is only in some such limited sense as this that the statement of the Triads can be true. We know positively that the Lloegrians of the south were reduced only by strenuous warfare, renewed campaign after campaign, and generation after generation, for hundreds of years. So, too, the natives of the north (who, however, do not seem to have been Lloegrians at all, but Brythons or Cymry), protracted their resistance, with various vicissitudes of defeat, subjection, and revolt, from the battle of Cattraeth to that of Dunmailraise. With regard to Essex and Mercia as far west as the Severn, Ave have so little information that it is quite possible that in certain tracts small British states, or Romano-British towns, may have allied themselves with the invaders, and thus for a time deferred their own destruction, though we cannot positively say that it was so. That anything of the kind occurred in East Anglia, unless on the smallest scale, I entirely disbelieve: the physical character- istics, the local names, the great number and moderate size of the hundreds and parishes, which justify an estimate of the free population at 50,000 or more, and the neighbourhood to the continental homes of the invaders, all tend to show that this region was very densely settled by a population almost exclusively Teutonic. Of Essex I have spoken

* The retention of the name Carnavy by the peasantry, as already mentioned, for a district in central Warwickshire should be remembered.

THE RACES OF BRITAIN. 65

already: its people, except near the coast, are more usually dark-haired, I think, than those of Kent, Norfolk, and Suffolk. That some compact may have been made between the invaders of Mercia and the southern portion of the Coritanian tribe, is rendered less improbable by the marked physical difference, already mentioned as noted by the late Professor Phillips as well as by myself, between the people of Leicester- shire and those of the counties to the north, although Leicestershire was very densely colonised by the Scandinavians in later times. Pal- grave conjectured that some of the Mercian counties might represent small British states, which, after holding out for a time, might by treaty, or even by royal marriages, have passed under the Saxon yoke. Possibly something of the kind may have happened to Rutland, or to Huntingdonshire (which is probably enough the South Gyrwa of earlier times), but I do not think we can admit the probability of the existence of any considerable British state in Mercia after the sixth century. We may conjecture that Mercia was built up out of a great number of small states or chieftainships, among which no one or two greatly preponderated for a generation or two after the beginning of Teu- tonic colonisation. The names of the marks, as investigated by Kemble, render it probable that this colonisation was naturally subsequent to that of the kingdoms on the coast, and much less dense, and this last fact is confirmed by certain statements of Bede's and others in the docu- ment in Sir Henry Spelman's glossary (article Hida)* as to the number of hydes in a number of districts, whose names are for the most part no longer recognisable with certainty, but some of which certainly, and most of which probably, did ultimately coalesce into the great Mercian kingdom. Thus, while Kent had 15,000 hydes, North and South Mercia had together but 12,000, and Hwicca is said to have contained only 600, only half the number of the Isle of Wight. This last statement is so extraordinary as to arouse one's suspicion ; however, Worcestershire contains but five hundreds at the present day, and Staffordshire the same number, while Sussex has 72.!

On the whole, the following is the theory of the early history of Mercia which most commends itself to me. About the close of the fifth century, a number of incoherent British tribes and decaying Romano- British towns, without any recognised head. This state of things first disturbed by the Anglian conquest of Lindisse, and the ravages of Cerdic about Oxfordshire. The country gradually penetrated by Saxon settlers from the already consolidated states of the south-east, and Anglians from East Anglia or Deira, or from the continent, who settle down in small bodies, sometimes enslaving or expelling a previous population, but sometimes peacefully reclaiming land from the waste

* Kemble i., 81. t Sussex is said to have contained 7,000 hydes ; if the number of hundreds was about the same as at present, there must have been a hundred hydes to each of them, i.e., probably a hundred free holdings, of, say, 30 acres each, or 60 at most.

6

66 THE RACES OF BRITAIN.

or the forest. Wars and alliances take place between the natives and the new-comers, generally on a small scale. In some districts the latter come thus to form merely a military aristocracy, but in the country north of the Trent, as far westwards at least as the Staffordshire and Derbyshire mountains and Cannock Chase, the whole population is destroyed, expelled, or enslaved, either by Ethelfrid Fleisawr of North- umbria, on his march towards Chester, or by an extension up the Trent valley of the Gainas and Lindiswaras. Thus a compact Anglian state, at first dependent on Deira, is formed, which in process of time, by conquest, agreement, or even marriage, absorbs the other Saxon states, and the remaining British ones, if any.

The philological evidence on the subject under discussion may be considered under two heads. The first is that of local names ; the second that of current language.

The local names of England have been investigated in detail by Leo and Isaac Taylor ; and Kemble has made some very valuable contribu- tions to our knowledge in this department.

Taylor, "in order to exhibit," as he says, "the comparative amount of the Celtic, the Saxon, and the Danish element of population in various portions of Britain," made an analysis of the names of hamlets, hills, woods, valleys, &c, in five counties, with the following result :

PERCENTAGE OF NAMES.

From the Suffolk. Surrey. Devon. Cornwall. Monmouth.

Celtic 2 8 32 80 76

Anglo-Saxon go 91 65 20 24

Norse 8 1 3 o o

These figures, however, cannot be taken as really exhibiting the proportions of the elements of population ; nor do I suppose that Mr. Taylor meant to claim for them anything more than an approximate and comparative value. For example, it is pretty certain that the Saxon element in the blood of Devon is nothing like 65 per cent., if, indeed, it be the half of 65 ; but it may very well be that the Saxon blood in Corn- wall bears to the Saxon blood in Devon some such proportion as 20 to 65. Again, Suffolk has fewer Celtic names than any other county on the list, and we may very fairly conclude that it has also less Celtic blood ; but every other line of argument tends to indicate that even in Suffolk the proportion of Celtic blood must be much greater than 2 per cent.

A somewhat similar investigation of the place-names in the south of Scotland yielded me the following results :

PERCENTAGE OF NAMES.

Berwick, Rox., Dumf., Lank., Renf., Ayr,

From the Selk., Peeb. Lothian. Stirling. Galloway.

Celtic 11 24 39 49

Doubtful 17 10 7 10

Saxon or Norse 72 66 54 41

THE RACES OF BRITAIN. 67

Here the Teutonic element in Galloway and the west is very much overstated, but whether it is much so in the first or Tweeddale group of counties is matter of doubt.

Many circumstances obviously affect the value of this kind of evidence. Thus, objects of great magnitude, or visible from afar, and therefore likely to be named and spoken of in the intercourse of two neighbouring races, such as rivers* and mountains, usually retain their original names through all changes of race and language. When a conquest is gradual, or very limited in extent, so as to be merely a small advanceof the frontier, comparatively few names will be changed, because, as a rule, the old names will have already become familiar to the conquerors. But with lapse of time the old names tend to die out, and to be replaced by new ones which are significant in the minds of the people, and which often nearly resemble the old ones in sound. And as all local names are at first significant, animo imponentis, all castles, houses and hamlets erected subsequently to the introduction of a new language will almost certainly be named in accordance with it. Suffolk was thoroughly English in speech more than a thousand years ago, but I have no doubt that at that period the proportion of Celtic local names there was very much greater than 2 per cent.

Where the invaders and the invaded are nearly equal in numbers, it may perhaps depend on national peculiarities of character and civilisation whether the former or the latter leave the more numerous traces of their presence in the local names. The Danes and Norsemen, for example, seem to have had a much greater power of giving local names than of transmitting their language. Both in Normandy and in the Hebrides they have left their mark in multitudes of local names ; but in the former their speech faded out in a generation or two, and in the latter it was not much more enduring. On the other hand, the colonisation of Ulster by the English and Scotch, though it involved an enormous change in the nationality of the population, and an entire change of language, had very little effect on the local names.

On the whole, the evidence of local names seems to favour the opinion already expressed, that the population in a great part of the East of England (and I may add the South-Eastern Lowlands, from the Cheviots almost to the Forth) have more Saxon than Celtic blood in their veins, and that a great part of the churls must have been Saxons. Of par- ticular districts, in which I have supposed the natives to have remained in large proportion, the Weald and the borders of Romney Marsh retain a number of Celtic or other ancient names, as Lympne, Appledore, Appledram, Glvnde, Rusper, Findon, and numerous Combes. The names in the Fenland are mostly modern ; but some of the more ancient appear to be Celtic : the most notable is a Gaelic one, Wiskin, i.e. water- island. In Cumberland, the nature and distribution of local names support in a mark worthy manner the views I shall presently put for- * Mr. Taylor has judiciously excluded river-names from his table.

6*

68 THE RACES OF BRITAIN.

ward. Many Celtic names, Gaelic as well as Cymric, remain, especially of natural objects on a large scale, as rivers and mountains ; but others have only Norse names, and considerable tracts occur in which the latter exclusively prevail. These facts accord with the hypothesis of the existence of a scanty and scattered British population, among and between whose occupations strong Norse colonies gradually settled down .

The investigation of the evidence of current language and provincial dialects is of considerable importance for our purpose, but that import- ance has perhaps been overrated by most of those who had entered on the quest. It does, however, tell us much, and would doubtless tell us more if more thoroughly worked out ; but the field of labour is wide, and the labour itself difficult, and the fruit often ambiguous. The rude old theory was, that language being the best evidence of national kin- ship and descent, and the English language being Teutonic, the English people must be Teutonic too. The history of the Cornish language furnishes a sufficient answer to the former of the above premises. The second one has seldom been doubted, until of late some have endea- voured to dispute it, on the ground that Teutonic words do not form half the contents of the English dictionary. There is no doubt, of course, that the current opinion on this point is the correct one. The commonest and most important and necessary of our words, particularly among the verbs, are Teutonic ; so are most of our scanty grammatical forms and rules ; and even in pronunciation (though this, too, has lately been disputed) those who have had opportunities of hearing Frisians or SchleswTigers or Jutlanders in conversation, and who have also listened to Welshmen similarly engaged, will surely agree with me that the English in general follow in intonation and cadence what may be called the paternal rather than the maternal side.

Some of the points urged by philo-Celtic writers are pretty certainly baseless ; e.g., the supposed Celtic feminine termination -ess, which was really brought from France by the Normans, and was not a Cymric legacy. A genuine remnant of the old British (Lloegrian) pronunciation may be found in a quarter where it might well be looked for, namely, in Devonshire, where, in travelling westwards, we encounter a people long thoroughly Anglicised, but some of whom seem to have spoken Cornish down to Elizabeth's time. Here, or more exactly on this side of the Devonshire border, but beyond the Parret and the Axe, the sound of the French u begins to be heard. The French probably inherit it from the Gauls, whose kindred, the Welsh, retain it, though they do not use it much. I do not think it is known to the Frisians.

Mr. Pike* quotes from William Edwards the statement that, in

Mezzofanti's opinion, the extreme irregularity of English pronunciation

is traceable to