VIII. ETHNOLOGICA.

ON THE SUBJECTIVITY OF CERTAIN CLASSES IN ETHNOLOGY.

FROM THE PHILOSOPHICAL MAGAZINE FOR MAY 1853.

To the investigator who believes in the unity of the human species, whether he be a proper ethnologist, or a zoologist in the more current signification of the term, the phænomena exhibited by the numerous families of mankind supply ninetenths of the data for that part of natural history which deals with varieties as subordinate to, and as different from, species. The history of domestic animals in comprehensiveness and complexity yields to the history of the domesticator. Compare upon this point such a work as G. Cuvier's on the Races of Dogs, with Dr. Prichard's Natural History of Man. The mere difference in bulk of volume is a rough measure of the difference in the magnitude of the subjects. Even if the dog were as ubiquitous as man, and consequently as much exposed to the influence of latitude, and altitude, there would still be wanting to the evolution of canine varieties the manifold and multiform influences of civilization. The name of these is legion; whilst the extent to which they rival the more material agencies of climate and nutrition is getting, day by day, more generally admitted by the best and most competent inquirers. Forms as extreme as any that can be found within the pale of the same species are to be found within that of the species Homo. Transitions as gradual as those between any varieties elsewhere are also to be found. In summing up the value of the data supplied by man towards the natural history of varieties, it may be said that they are those of a species which has its geographical distribution everywhere and a moral as well as a physical series of characteristics. Surely, if the question under notice be a question that must be studied inductively, Man gives us the field for our induction.

Before I come to the special point of the present notice and to the explanation of its somewhat enigmatical heading, I must further define the sort of doctrine embodied in what I have called the belief of the unity of our species. I do not call the upholder of the developmental doctrine a believer of this kind. His views—whether right or wrong—are at variance with the current ideas attached to the word species. Neither do I identify with the recognition of single species the hypothesis of a multiplicity of protoplasts, so long as they are distributed over several geographical centres. The essential element to the idea of a single species is a single geographical centre. For this, the simplest form of the protoplast community is a single pair.

All this is mere definition and illustration. The doctrine itself may be either right or wrong. I pass no opinion upon it. I assume it for the present; since I wish to criticize certain terms and doctrines which have grown up under the belief in it, and to show, that, from one point of view, they are faulty, from another, legitimate.

It will simplify the question if we lay out of our account altogether the islands of the earth's surface, limiting ourselves to the populations of the continent. Here the area is continuous, and we cannot but suppose the stream of population by which its several portions were occupied to have been continuous also. In this case a population spreads from a centre like circles on a still piece of water. Now, if so, all changes must have been gradual, and all extreme forms must have passed into each other by means of a series of transitional ones.

It is clear that such forms, when submitted to arrangement and classification, will not come out in any definite and wellmarked groups, like the groups that constitute what is currently called species. On the contrary, they will run into each other, with equivocal points of contact, and indistinct lines of demarcation; so that discrimination will be difficult, if not impracticable. If practicable, however, it will be effected by having recourse to certain typical forms, around which such as approximate most closely can most accurately and conveniently be grouped. When this is done, the more distant outliers will be distributed over the debateable ground of an equivocal frontier. To recapitulate: varieties as opposed to species imply transitional forms, whilst transitional forms preclude definite lines of demarcation.

Yet what is the actual classification of the varieties of mankind, and what is the current nomenclature? To say the least, it is very like that of the species of a genus. Blumenbach's Mongolians, Blumenbach's Caucasians, Blumenbach's Æthiopians,—where do we find the patent evidence that these are the names of varieties rather than species? Nowhere. The practical proof of a clear consciousness on the part of a writer that he is classifying varieties rather than species, is the care he takes to guard his reader against mistaking the one for the other, and the attention he bestows on the transition from one type to another. Who has ever spent much ethnology on this? So far from learned men having done so, they have introduced a new and lax term—race. This means something which is neither a variety nor yet a species—a tertium quid. In what way it differs from the other denomination has yet to be shown.

Now if it be believed (and this belief is assumed) that the varieties of mankind are varieties of a species only, and if it cannot be denied that the nomenclature and classification of ethnologists is the nomenclature and classification of men investigating the species of a genus, what is to be done? Are species to be admitted, or is the nomenclature to be abandoned? The present remarks are made with the view of showing that the adoption of either alternative would be inconsiderate, and that the existing nomenclature, even when founded upon the assumption of broad and trenchant lines of demarcation between varieties which (ex vi termini) ought to graduate into each other, is far from being indefensible.

Man conquers man, and occupant displaces occupant on the earth's surface. By this means forms and varieties which once existed become extinct. The more this extinction takes place, the greater is the obliteration of those transitional and intermediate forms which connect extreme types; and the greater this obliteration, the stronger the lines of demarcation between geographically contiguous families. Hence a variational modification of a group of individuals simulates a difference of species; forms which were once wide apart being brought into juxtaposition by means of the annihilation of the intervening transitions. Hence what we of the nineteenth century,—ethnologists, politicians, naturalists, and the like—behold in the way of groups, classes, tribes, families, or what not, is beholden to a great extent under the guise of species; although it may not be so in reality, and although it might not have been so had we been witnesses to that earlier condition of things when one variety graduated into another and the integrity of the chain of likeness was intact. This explains the term subjectivity. A group is sharply defined simply because we know it in its state of definitude; a state of definitude which has been brought about by the displacement and obliteration of transitional forms.

The geographical distribution of the different ethnological divisions supplies a full and sufficient confirmation of this view. I say "full and sufficient," because it cannot be said that all our groups are subjective, all brought about by displacement and obliteration. Some are due to simple isolation; and this is the reason why the question was simplified by the omission of all the insular populations. As a general rule, however, the more definite the class, the greater the displacement; displacement which we sometimes know to have taken place on historical evidence, and displacement which we sometimes have to infer. In thus inferring it, the language is the chief test. The greater the area over which it is spoken with but little or no variation of dialect, the more recent the extension of the population that speaks it. Such, at least, is the primâ facie view.

A brief sketch of the chief details that thus verify the position of the text is all that can now be given.

1. The populations of South-eastern Asia, Mongol in physiognomy and monosyllabic in speech, have always been considered to form a large and natural, though not always a primary, group. Two-thirds of its area, and the whole of its frontier north of the Himalayas, is formed by the Chinese and Tibetans alone. These differ considerably from each other, but more from the Turks, Mongols, and Tongusians around. In the mountainous parts of the Assam frontier and the Burmese empire, each valley has its separate dialect. Yet these graduate into each other.

2. Central Asia and Siberia are occupied by four great groups, the populations allied to the Turk, the populations allied to the Mongol, the populations allied to the Mantshu, and the populations allied to the Finns. These are pretty definitely distinguished from each other, as well as from the Chinese and Tibetans. They cover a vast area, an area, which, either from history or inference, we are certain is far wider at present than it was originally. They have encroached on each and all of the populations around, till they meet with families equally encroaching in the direction of China and Tibet. This it is that makes the families which are called Turanian and Monosyllabic natural groups. They are cut off, more or less, from each other and from other populations by the displacement of groups originally more or less transitional. The typical populations of the centre spread themselves at the expense of the sub-typicals of the periphery until the extremes meet.

3. The circumpolar populations supply similar illustrations. Beginning with Scandinavia, the Lap stands in remarkable contrast with the Norwegian of Norway, and the Swede of Sweden. Why is this? Because the Northman represents a population originally German,—a population which, however much it may have graduated into the type of the most southern congeners of the Lap, is now brought into contact with a very different member of that stock.

4. This phænomenon repeats itself in the arctic portions of America, where the Algonkin and Loucheux Indians (Indians of the true American type) come in geographical contact, and in physiological contrast, with the Eskimo. Consequently along the Loucheux and Algonkin frontiers the line of demarcation between the Eskimo and the Red Indian (currently so-called) is abrupt and trenchant. Elsewhere, as along the coast of the Pacific, the two classes of population graduate into each other.

5. The African family is eminently isolated. It is, however, just along the point of contact between Africa and Asia that the displacements have been at a maximum. The three vast families of the Berbers, the Arabs and the Persians, cannot but have obliterated something (perhaps much) in the way of transition.

6. The Bushmen and Hottentots are other instances of extreme contrast, i. e. when compared with the Amakosah Caffres. Yet the contrast is only at its height in those parts where the proof of Caffre encroachment is clearest. In the parts east of Wallfisch Bay—traversed by Mr. Galton—the lines of difference are much less striking.

Such are some of the instances that illustrate what may be called the "subjectivity of ethnological groups,"—a term which greatly helps to reconcile two apparently conflicting habits, viz. that of thinking with the advocates of the unity of the human species, and employing the nomenclature of their opponents.

GENERAL PRINCIPLES OF PHILOLOGICAL CLASSIFICATION AND THE VALUE OF GROUPS,

WITH PARTICULAR REFERENCE TO THE LANGUAGES OF THE INDO-EUROPEAN CLASS.

READ BEFORE THE ETHNOLOGICAL SOCIETY, 28TH FEBRUARY 1849.

In respect to the languages of the Indo-European class, it is considered that the most important questions connected with their systematic arrangement, and viewed with reference to the extent to which they engage the attention of the present writers of philology, are the three following:—

1. The question of the Fundamental Elements of certain Languages.—The particular example of an investigation of this kind is to be found in the discussion concerning the extent to which it is a language akin to the Sanskrit, or a language akin to the Tamul, which forms the basis of certain dialects of middle and even northern India. In this is involved the question as to the relative value of grammatical and glossarial coincidences.

2. The question of the Independent or Subordinate Character of certain Groups.—Under this head comes the investigation, as to whether the Slavonic and Lithuanic tongues form separate groups, in the way that the Slavonic and Gothic tongues form separate groups, or whether they are each members of some higher group. The same inquiry applies to the languages (real or supposed) derived from the Zend, and the languages (real or supposed) derived from the Sanskrit.

3. The question of Extension and Addition.—It is to this that the forthcoming observations are limited.

Taking as the centre of a group, those forms of speech which have been recognised as Indo-European (or Indo-Germanic), from the first recognition of the group itself, we find the languages derived from the ancient Sanskrit, the languages derived from the ancient Persian, the languages of Greece and Rome, the Slavonic and Lithuanic languages, and the languages of the Gothic stock; Scandinavian, as well as Germanic. The affinity between any two of these groups has currently been considered to represent the affinity between them all at large.

The way in which the class under which these divisions were contained, as subordinate groups, has received either addition or extension, is a point of philological history, which can only be briefly noticed; previous to which a difference of meaning between the words addition and extension should be explained.

To draw an illustration from the common ties of relationship, as between man and man, it is clear that a family may be enlarged in two ways.

a. A brother, or a cousin, may be discovered, of which the existence was previously unknown. Herein the family is enlarged, or increased, by the real addition of a new member, in a recognised degree of relationship.

b. A degree of relationship previously unrecognised may be recognised, i. e., a family wherein it was previously considered that a second-cousinship was as much as could be admitted within its pale, may incorporate third, fourth, or fifth cousins. Here the family is enlarged, or increased, by a verbal extension of the term.

Now it is believed that the distinction between increase by the way of real addition, and increase by the way of verbal extension, has not been sufficiently attended to. Yet, that it should be more closely attended to, is evident; since, in mistaking a verbal increase for a real one, the whole end and aim of classification is overlooked.

I. The Celtic.—The publication of Dr. Prichard's Eastern Origin of the Celtic Nations, in 1831, supplied philologists with the most definite addition that has, perhaps, yet been made to ethnographical philology.

Ever since then, the Celtic has been considered to be Indo-European. Indeed its position in the same group with the Iranian, Classical, Slavono-Lithuanic, and Gothic tongues, supplied the reason for substituting the term Indo-European for the previous one Indo-Germanic.

2. Since the fixation of the Celtic, it has been considered that the Armenian is Indo-European. Perhaps the wellknown affinity between the Armenian and Phrygian languages directed philologists to a comparison between the Armenian and Greek. Müller, in his Dorians, points out the inflexion of the Armenian verb-substantive.

3. Since the fixation of the Celtic, it has been considered that the old Etruscan is Indo-European.

4. Since the fixation of the Celtic, it has been considered that the Albanian is Indo-European.

5. Since the fixation of the Celtic, Indo-European elements have been indicated in the Malay.

6. Since the fixation of the Celtic, Indo-European elements have been indicated in the Laplandic.

7. Since the fixation of the Celtic, it has been considered that the Ossetic is Indo-European.

8. Since the consideration of the Ossetic as Indo-European, the Georgian has been considered as Indo-European likewise.

Now the criticism of the theory which makes the Georgian to be Indo-European, is closely connected with the criticism of the theory which makes the Ossetic and the Malay to be Polynesian; and this the writer reserves for a separate paper. All that he does at present is to express his opinion, that if any of the seven last-named languages are Indo-European, they are Indo-European not by real addition, in the way of recognised relationship, but by a verbal extension of the power of the term Indo-European. He also believes that this is the view which is taken, more or less consciously or unconsciously, by the different authors of the different classifications themselves. If he be wrong in this notion, he is at issue with them as to a matter of fact; since, admitting some affinity on the part of the languages in question, he denies that it is that affinity which connects the Greek and German, the Latin and Lithuanian.

On the other hand, if he rightly imagine that they are considered as Indo-European on the strength of some other affinity, wider and more distant than that which connects the Greek with the German, or the Latin with the Lithuanic, he regrets that such an extension of a term should have been made without an exposition of the principles that suggested it, or the facts by which it is supported; principles and facts which, when examined by himself, have convinced him that most of the later movements in this department of ethnographical philology, have been movements in the wrong direction.

There are two principles upon which languages may be classified.

According to the first, we take two or more languages as we find them, ascertain certain of their characteristics, and then inquire how far these characteristics coincide.

Two or more languages thus taken agree in having a large per-centage of words in common, or a large per-centage of grammatical inflexions; in which case they would agree in certain positive characters. On the other hand, two or more such languages agree in the negative fact of having a small and scanty vocabulary, and an inflexional system equally limited; whilst, again, the scantiness of inflexion may arise from one of two causes. It may arise from the fact of inflexions having never been developed at all, or it may arise from inflexions having been lost subsequent to a full development of the same. In all such cases as these, the principle of classification would be founded upon the extent to which languages agreed or differed in certain external characteristics; and it would be the principle upon which the mineralogist classifies minerals. It is not worth while to recommend the adoption of the particular term mineralogical, although mineralogy is the science that best illustrates the distinction. It is sufficient to state, that in the principle here indicated, there is no notion of descent.

It is well known that in ethnographical philology (indeed in ethnology at large) the mineralogical principle is not recognised; and that the principle that is recognised is what may be called the historical principle. Languages are arranged in the same class, not because they agree in having a copious grammar or scanty grammar, but because they are descended (or are supposed to be descended) from some common stock; whilst similarity of grammatical structure, and glossarial identity are recognised as elements of classification only so far as they are evidence of such community of origin. Just as two brothers will always be two brothers, notwithstanding differences of stature, feature, and disposition, so will two languages which have parted from the common stock within the same decennium, be more closely allied to each other, at any time and at all times, than two languages separated within the same century; and two languages separated within the same century, will always be more cognate than two within the same millennium. This will be the case irrespective of any amount of subsequent similarity or dissimilarity.

Indeed, for the purposes of ethnology, the phenomena of subsequent similarity or dissimilarity are of subordinate importance. Why they are so, is involved in the question as to the rate of change in language. Of two tongues separated at the same time from a common stock, one may change rapidly, the other slowly; and, hence, a dissimilar physiognomy at the end of a given period. If the English of Australia were to change rapidly in one direction, and the English of America in another, great as would be the difference resulting from such changes, their ethnological relation would be the same. They would still have the same affiliation with the same mother-tongue, dating from nearly the same epoch.

In ethnological philology, as in natural history, descent is the paramount fact; and without asking how far the value thus given to it is liable to be refined on, we leave it, in each science, as we find it, until some future investigator shall have shewn that either for a pair of animals not descended from a common stock, or for a pair of languages not originating from the same mother-tongue, a greater number of general propositions can be predicated than is the case with the two most dissimilar instances of either an animal or a language derived from a common origin.

Languages are allied just in proportion as they were separated from the same language at the same epoch.

The same epoch.—The word epoch is an equivocal word, and it is used designedly because it is so. Its two meanings require to be indicated, and, then, it will be necessary to ask which of them is to be adopted here.

The epoch, as a period in the duration of a language, may be simply chronological, or it may be philological, properly so called.

The space of ten, twenty, a hundred, or a thousand years, is a strictly chronological epoch. The first fifty years after the Norman conquest is an epoch in the history of the English language; so is the reign of Henry the Third, or the Protectorship of Oliver Cromwell. A definite period of this sort is an epoch in language, just as the term of twenty or thirty years is an epoch in the life of a man.

On the other hand, a period that, chronologically speaking, is indefinite, may be an epoch. The interval between one change and an other, whether long or short, is an epoch. The duration of English like the English of Chaucer, is an epoch in the history of the English language; and so is the duration of English like the English of the Bible translation. For such epochs there are no fixed periods. With a language that changes rapidly they are short; with a language that changes slowly they are long.

Now, in which of these two meanings should the word be used in ethnographical philology? The answer to the question is supplied by the circumstances of the case, rather than by any abstract propriety. We cannot give it the first meaning, even if we wish to do so. To say in what year of the duration of a common mother-tongue the Greek separated from the stock that was common to it and to the Latin is an impossibility; indeed, if it could be answered at once, it would be a question of simple history, not an inference from ethnology: since ethnology, with its palæontological reasoning from effect to cause, speaks only where history, with its direct testimony, is silent.

We cannot, then, in ethnological reasoning, get at the precise year in which any one or two languages separated from a common stock, so as to say that this separated so long before the other.

The order, however, of separation we can get at; since we can infer it from the condition of the mother-tongue at the time of such separation; this condition being denoted by the condition of the derived language.

Hence the philological epoch is an approximation to the chronological epoch, and as it is the nearest approximation that can possibly be attained, it is practically identical with it, so that the enunciation of the principle at which we wish to arrive may change its wording, and now stand as follows,—Languages are allied, just in proportion as they were separated from the same language in the same stage.

Now, if there be a certain number of well-marked forms (say three) of development, and if the one of these coincide with an early period in the history of language, another with a later one, and the third with a period later still, we have three epochs wherein we may fix the date of the separation of the different languages from their different parent-stocks; and these epochs are natural, just in proportion as the forms that characterise them are natural.

Again, if each epoch fall into minor and subordinate periods, characterised by the changes and modifications of the then generally characteristic forms, we have the basis for subordinate groups and a more minute classification.

It is not saying too much to say that all this is no hypothesis, but a reality. There are real distinctions of characteristic forms corresponding with real stages of development; and the number of these is three; besides which, one, at least, of the three great stages falls into divisions and subdivisions.

1. The stage anterior to the evolution of inflexion.—Here each word has but one form, and relation is expressed by mere juxtaposition, with or without the superaddition of a change of accent. The tendencies of this stage are to combine words in the way of composition, but not to go further. Every word retains, throughout, its separate substantive character, and has a meaning independent of its juxtaposition with the words with which it combines.

2. The stage wherein inflexions are developed.—Here, words originally separate, and afterwards placed in juxtaposition with others, as elements of a compound term, so far change in form, or so far lose their separate signification, as to pass for adjuncts, either prefixed or postfixed to the main word. What was once a word is now the part of a word, and what was once Composition is now Derivation, certain sorts of Derivation being called Inflexions, and certain Inflexions being called Declensions or Conjugations, as the case may be.

3. The stage wherein inflexions become lost, and are replaced by separate words.—Here case-endings, like the i in patr-i, are replaced by prepositions (in some cases by postpositions), like the to in to father; and personal endings, like the o in voc-o, are replaced by pronouns, like the I in I call.

Of the first of these stages, the Chinese is the language which affords the most typical specimen that can be found in the present late date of languages—late, considering that we are looking for a sample of its earliest forms.

Of the last of these stages the English of the year 1849 affords the most typical specimen that can be found in the present early date of language—early, considering that we are looking for a sample of its latest forms.

Of the second of these stages we must take two languages as the samples.

1. The Greek.—Here we have the inflexional character in its most perfect form; i. e., the existence, as separate words, of those sounds and syllables that form inflexions is at its maximum of concealment; i. e., their amalgamation with the primary word (the essence of inflexion) is most perfect.

2. The Circassian, Coptic, or Turkish.—In one of these (it is difficult to say which) the existence as separate words of those sounds and syllables which form inflexions, is at its minimum of concealment; i. e., their amalgamation with the primary word (the essence of inflexion) being most imperfect.

This classification is, necessarily, liable to an element of confusion common to all classifications where the evidence is not exactly of the sort required by the nature of the question. The nature of the question here dealt with requires the evidence of the historical kind, i. e., direct testimony. The only evidence, however, we can get at is indirect and inferential. This engenders the following difficulty. The newest language of (say) the languages of the secondary formation may be nearer in chronology, to the oldest language of the third, than to the first formed language of its own class. Indeed, unless we assume the suspension of all change for long epochs, and that those coincide with the periods at which certain languages are given off from their parent stocks, such must be the case.

Now, although this is a difficulty, it is no greater difficulty than the geologists must put up with. With them also there are the phenomena of transition, and such phenomena engender unavoidable complications. They do so, however, without overthrowing the principles of their classification.

The position of a language in respect to its stage of development is one thing,—the position in respect to its allied tongues another.

Two languages may be in the same stage (and, as such, agree), yet be very distant from each other in respect to affiliation or affinity. Stage for stage the French is more closely connected with the English, than the English with the Mœso-Gothic. In the way of affiliation, the converse is the case.

Languages are allied (or, what is the same thing, bear evidence of their alliance), according to the number of forms that they have in common; since (subject to one exception) these common forms must have been taken from the common mother-tongue.

Two languages separated from the common mother-tongue, subsequent to the evolution of (say) a form for the dative case, are more allied than two languages similarly separated anterior to such an evolution.

Subject to one exception. This means, that it is possible that two languages may appear under certain circumstances more allied than they really are, and vice versâ.

They may appear more allied than they really are, when, after separating from the common mother-tongue during the ante-inflexional stage, they develop their inflexions on the same principle, although independently. This case is more possible than proved.

They may appear less allied than they really are, when, although separated from the common mother-tongue after the evolution of a considerable amount of inflexion, each taking with it those inflexions, the one may retain them, whilst the other loses them in toto. This case also is more possible than proved.

Each of these cases involves a complex question in philology:—the one the phenomena connected with the rate of change; the other the uniformity of independent processes.

These questions are likely to affect future researches more than they have affected the researches hitherto established. Another question has affected the researches hitherto established more than it is likely to affect future ones. This is the question as to the fundamental unity, or non-unity of language. Upon this the present writer has expressed an opinion elsewhere. At present he suggests that the more the general unity of the human language is admitted, the clearer will be the way for those who work at the details of the different affiliations. As long as it is an open question, whether one class of languages be wholly unconnected with others, any connection engenders an inclination to arrange it under the group previously recognised. I believe that this determined the position of the Celtic in the Indo-European group. I have great doubts whether if some affinity had been recognised from the beginning, it would even have stood where it now does. The question, when Dr. Prichard undertook his investigations, was not so much whether the Celtic was in the exact ratio to any or all of the then recognised European languages in which they were to each other, but whether it was in any relation at all. This being proved, it fell into the class at once.

The present writer believes that the Celtic tongues were separated from their mother-tongue at a comparatively early period of the second stage; i. e., when but few inflexions had been evolved; whilst the Classic, Gothic, Lithuano-Slavonic (Sarmatian), and Indo-Persian (Iranian) were separated at comparatively late periods of the same stage, i. e., when many inflexions had been evolved.

Hence he believes that, in order to admit the Celtic, the meaning of the term Indo-European was extended.

Regretting this (at the same time admitting that the Celtic tongue is more Indo-European than any thing else), he believes that it is too late to go back to the older and more restricted use of the term; and suggests (as the next best change), the propriety of considering the Indo-European class as divided into two divisions, the older containing the Celtic, the newer containing the Iranian, Classical, Sarmatian, and Gothic tongues. All further extensions of the term he believes to be prejudicial to future philology; believing also that all supposed additions to the Indo-European class have (with the exception, perhaps, of the Armenian) involved such farther extension.

TRACES OF A BILINGUAL TOWN IN ENGLAND.

READ AT THE MEETING OF THE BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE 1853.

It is well-known that the termination -by as the name of a village or town is a sign of Danish occupancy. At the present time it means town in Scandinavia; and Christiania or Copenhagen is called By, or Byen, = the town, capital, or metropolis. The English form is -ton. When an Angle said Newton, a Dane said Newby. The distribution of the forms in -by has already commanded much attention; so that it is not the intention of the present writer to say much about it.

Along, however, with this form go others: e. g.

The EnglishShipbecomes in DanishSkipas inSkipton

——Fish——Fisk——Fiskerton

——Worm——Orm——Ormsby

——Church——Kirk——Ormskirk

&c. &c.

In like manner the Roman castra becomes—

In English chester or cester, in Danish caster and caistor. Contrast the forms Tadcaster, Lancaster &c. with Chester, or Bicester and this difference becomes apparent.

Now the river Ouse in the parts about Wansford separates the counties of Huntingdon and Northampton—in the former of which no place ending in -by is to be found, and all the castra are chester; as Godmanchester. In Northamptonshire, on the other hand, the Danish forms in -by are common, and the castra are caistor, or caster. All the Danish is on one side. Nothing is Danish on the other. The river has every appearance of having formed a frontier. On it lay the Roman station of Durobrivis—with, probably, castra on each side. At any rate, there are, at the present moment, two villages wherein that term appears. On the Huntingdon side is the village of Chesterton (English). On the Northampton side is that of Caistor (Danish).

ON THE ETHNOLOGICAL POSITION OF CERTAIN TRIBES ON THE GARROW HILLS.

READ AT THE MEETING OF THE BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE HELD AT YORK 1844.

The affinities of the Garrow language, a language which Klaproth in his Asia Polyglotta leaves unplaced, are with the Tibetan.

The bearings of this will be found in the next notice.

NOTE (1859).

This was written before I had seen Brown's Tables—wherein the affinity is virtually, though not directly affirmed.

ON THE TRANSITION BETWEEN THE TIBETAN AND INDIAN FAMILIES IN RESPECT TO CONFORMATION.

BRITISH ASSOCIATION—BIRMINGHAM 1849.

The remarks of Mr. Hodgson on the Kooch, Bodo, and Dhimal, along with some of Dr. Bird's on the monosyllabic affinities of the Tamulian languages have an important bearing on this question. So have the accounts of the Chepang and Garo tribes. The phenomena are those of transition.

We have a practical instance of this in the doctrine laid down by Mr. Hodgson in his valuable monograph. In this, he makes the Bodo a Tamulian i. e. a member of the same family with the hill-tribes of India and the Dekhan; meaning thereby the aborigines of India, contrasted with the populations to which he ascribes the Sanskrit language and the Hindu physiognomy. In the Tamulian form there is "a somewhat lozenge contour, caused by the large cheek-bones"—"a broader flatter face"—"eyes less evenly crossing the face in their line of picture"—"beard deficient"—"with regard to the peculiar races of the latter" (i. e. the Tamulians) "it can only be safely said that the mountaineers exhibit the Mongolian type of mankind more distinctly than the lowlanders, and that they have, in general, a paler yellower hue than the latter, amongst whom there are some (individuals at least) who are nearly as black as negroes.—The Bodo are scarcely darker than the mountaineers above them—whom they resemble—only with all the physiognomical characteristics softened down.—The Kols have a similar cast of face."

This is the evidence of a competent observer to the fact of the Bodo &c. being, more or less, what is called Mongol; all the more valuable because he had not, then, recognized their language as monosyllabic. Meanwhile he never separates them from the Kols &c. but always connects the two. In other words, he gives us so much evidence to the fact of the Kols &c. being, more or less, Mongol also. But the Kols are the aborigines of India; whilst the Bodo are Tibetan.

NOTE (1859).

Recent researches have a tendency to make the Kols less Tamul and more Tibetan than they were held to be in 1849.

ON THE AFFINITIES OF THE LANGUAGES OF CAUCASUS WITH THE MONOSYLLABIC LANGUAGES.

READ AT THE MEETING OF THE BRITISH ASSOCIATION AT CAMBRIDGE 1845.

Taking the samples of the Georgian, Lesgian, Mizhdzhedzhi, and Circassian classes as we find them in the Asia Polyglotta and comparing them with the specimens of the monosyllabic languages in the same work, in Brown's Tables, and in Leyden's paper on the Indo-Chinese Languages, we find the following coincidences.[7]

English, sky

1. Circassian, whapeh, wuafe

2. Aka, aupa

Khamti, fa

English, sky

1. Absné, kaukh

Altekesek, hak

2. Akush, kaka

Burmese, kydukkhe

English, sky

1. Tshetshentsh, tulak

2. Koreng, talo

Khoibu, thullung

English, sun

1. Georgian, mse

Mingrelian, bsha

Suanic, mizh

2. Kuan-chua, zhi

Sianlo, suu

English, fire

1. Absné, mza

Circassian, mafa

2. Khamti, fai

Siam, fai

Aka, umma

Aber, eme

Burmese, mi

Karyen, me

Manipur, mai

Songphu, mai

Kapwi, &c., mai

English, day

1. Tshetshentsh, dini

Ingúsh, den

Kasikumuk, kini

2. Koreng, nin

Jili, tana

Singpho, sini

English, day

1. Andi, thyal

2. Garo, salo

English, moon

1. Georgian, twai=month

Suanic, twai

2. Moitay, ta

English, star

1. Kasikumuk, zuka

2. Garo, asake

Jili, sakan

Singpho, sagan

English, hill

1. Kasikumuk, suntu

2. Chinese, shan

English, earth

1. Absné, tshullah

Altekesek, tzula

2. Kapwi, talai

Khoibu, thalai

English, earth

1. Andi, zkhur

2. Mishimi, tari

English, earth

1. Dido, tshedo

2. Koreng, kadi

English, snow

1. Lesgian, asu

Circassian, uas

Abassian, asse

2. Chinese, siwe

English, salt

1. Lesgian[8](3), zam

2. Chinese, yan

English, salt

1. Kabutsh, tshea

Dido, zio

Kasikumuk, psu

Akush, dze

2. Tibetan, tsha

English, dust

1. Tshetshentsh, tshen

2. Chinese, tshin

English, sand

1. Avar, tshimig

2. Tibetan, bydzoma

English, sand

1. Circassian, pshakhoh

2. Chinese, sha

English, leaf

1. Tshetshentsh, ga

Ingush, ga

2. Chinese, ye

English, tree

1. Mizjeji, che

Circassian, dzeg

2. Chinese, shu

English, stone

1. Andi, hinzo

2. Siamese, hin

English, sea

1. Georgian, sgwa

2. Chinese, shuy=water

Tibet, çi=do

Môn, zhe=do

Ava, te=do

English, river

1. Anzukh, or kyare

Avar, hor, khor

2. Champhung, urai

English, river

1. Abassian, aji

2. Tibetan, tshavo

English, river

1. Altekesek, sedu

Absné, dzedu

2. Songphu, duidai

English, water

1. Kasikumuk, sin

Akush, shen

Kubitsh, tzun, sin

2. Singpho, ntsin

Jili, mchin

Manipur, ising

English, water

1. Absné, dzeh

2. Songphu, dui

Kapwi, tui

Tankhul, tu

English, water

1. Mizjeji, chi

2. Garo, chi

English, rain

1. Andi, za

Ingush, du

Abassian, kua

2. Chinese, yu

English, summer

1. Tushi, chko

Mizjeji, achke

2. Chinese, chia

English, winter

1. Anzukh, tlin

Andi, klinu

Kasikumuk, kintul

Akush, chani

Absné, gene

2. Tibetan, r gun

Chinese, tung

English, cow

1. Circassian, bsa

2. Tibetan, r shu

English, dog

1. Avar, choi

Andi, choi

Dido, gwai

Kubitsh, koy

Circassian, khhah

2. Chinese, keu

Tibetan, kyi

English, horse

1. Lesgian, tshu

Circassian, tshe, shu

2. Tibetan, r dda

English, bird

1. Avar, hedo

2. Tankhul, ata

English, bird

1. Andi, purtie

2. Abor, pettang

Aka, put'ah

English, fish

1. Avar, tshua

Circassian, bbzheh

2. Khamti, pa

Siamese, pla

Aka, ngay

Abor, engo

Burmese, nga

Karyen, nga

Singpho, nga

Songphu, kha

Mishimi, ta

Maram, khai

Luhuppa, khai

Tankhul, khi

Anam, khi

English, flesh

1. Kabutsh, kho

Abassian, zheh

2. Chinese, shou

Tibetan, zhsha

English, egg

1. Tshetshentsh, khua

2. Khamti, khai

Siamese, khai

English, egg

1. Kabutsh, tshemuza

2. Mishimi, mtiumaie

English, egg

1. Akush, dukhi

2. Garo, to`ka

English, son

1. Mizjeji, ua, woe

2. Tibetan, bu

English, hair

1. Kasikumuk, tshara

2. Jili, kara

Singpho, kara

English, hair

1. Avar, sab

Anzukh, sab

Tshari, sab

2. Burmese, shaben

Manipur, sam

Songpho (6), sam

English, hair

1. Tshetshentsh, kazeresh

2. Karyen, khosu

Tankhul, kosen

English, head

1. Georgian, tawi

Lazic, ti

Suanic, tchum

2. Chinese, teu, seu

Anam, tu ď u

Ava, kang (5)

English, head

1. Andi, mier, maœr

2. Assam, mur

English, head

1. Absné, kah, aka

Altekesek, zeka

2. Karen, kho

Manipur, kok

Taukhul, akao

English, mouth

1. Lesgian, kall

2. Chinese, keu

Anamese, kau

Tibetan, ka

English, mouth

1. Tushi, bak

2. Teina, pak

English, mouth

1. Georgian, piri

Mingrelian, pidehi

Suanic, pil

2. Ava, parat (4)

English, mouth

1. Kubitsh, mole

2. Khoibu, mur

Maring, mur

English, mouth

1. Andi, kol, tkol

Lesgian (3), kaal

2. Manipur, chil

English, eye

1. Andi, puni

2. Chinese, yan

English, ear

1. Avar, een, ain, en

Anzukh, in

Tshari, een, ein

Andi, kanka, andika

2. Burmese, na

Karen, naku

Singpho, na

Songphu, anhukon

Kapwi, kana

Koreng, kon

Maram, inkon

Champhung, khunu

Luhuppa, khana

Tankhul, akhana

Koibu, khana

English, tooth

1. Lesgian (3), sibi

Avar, zavi

Circassian, dzeh

2. Tibetan, so

Chinese, tshi

English, tongue

1. Circassian, bbse

Absné, ibs

2. Tibetan, rdzhe

Chinese, shi

English, foot

1. Kasikumuk, dzhan

2. Khamti, tin

English, foot

1. Mizjeji (3), kog, koeg

2. Manipur, khong

Tankhul, akho

English, foot

1. Andi, tsheka

Kubitsh, tag

Jili, takkhyai

2. Garo, jachok

English, foot

1. Georgian, pechi

2. Maplu, pokâ=leg

English, finger

1. Mingrelian, kiti

Moitay, khoit=hand

2. Play, kozu=do

English, hand

1. Georgian, chéli

Lazic, ieh

Mingrelian, ché

Suanic, shi

2. Chinese, sheu

English, hand

1. Andi, katshu

Kabutsh, koda

2. Khoibu, khut

Manipur, khut

English, blood

1. Absné, tsha, sha

Tshetshentsh, zi

Ingús, zi

2. Singpho, sai

Songpho, zyai

Kapwi, the

Maram, azyi

Champhung, azi

Luhuppa, ashi

Tankhul, asu

English, blood

1. Dido, é

2. Manipur, i

Koibu, hi

Maring, hi

English, blood

1. Tshetshentsh, yioh

Circassian, tlih

2. Chinese, chine

English, skin

1. Circassian, ffeh

2. Chinese, pi

English, skin

1. Dido, bik

2. Tibetan, shbagsbba

English, bone

1. Tshetshentsh, dyackt

Ingúsh, tekhh

Akúsh, likka

Tshari, rekka

2. Khamti, nuk

Siamese, kraduk

English, great

1. Georgian, didi

Mingrelian, didi

2. Canton, ta

Kuan-chua, ta, da

Tonkin, drai

Cochin-chinese, dai

Tibet, çe

Ava, kyi (5)

Play, du

Teina, to

English, bad

1. Mingrelian, moglach

Suanic, choya

2. Chinese, go gok

Môn, kah

Ava, makaung (4)

—— gye (2)

English, warm

1. Ingush, tau

2. Tibetan, dzho

English, blue

1. Mizjeji (3), siene

2. Chinese, zing

Tibetan, swongbba

English, yellow

1. Circassian, khozh

2. Abassian, kha

Chinese, chuang

English, green

1. Avar, ursheria

Anzukh, ordjin

Ingush, send

2. Tibetan, shjanggu

English, below

1. Georgian, kwewrt, kwerno

2. Ava, haukma

Yo, auk

Passuko, hoko

Kolaun, akoa

English, one

1. Lesgian, zo

Akush, za

Andi, sew

Dido, zis

Kasikumuk, zabá

Mizjeji, tza

Abassian, seka

2. Tibetan, dzig

English, three

1. Georgian, sami

Lazic, jum

Mingrelian, sami

Suanic, semi

2. Canton Chinese, sam

Kuanchua, san

Tonkin, tam

Tibetan, sum

Môn, sum

Ava, thaum

Siam, sum

English, four

1. Abassian, pshi ba

2. Tibetan, bshi

Chinese, szu

English, five

1. Georgian, chuthi

Lazic, chut

Mingrelian, chuthi

Suanic, wochu'si

2. Ava, yadu

English, six

1. Tshetshentsh, yatsh

Ingush, yatsh

Tushi, itsh

2. Tibetan, dzhug

English, nine

1. Circassian, bgu

2. Tibetan, rgu

Chinese, kieu

English, ten

1. Circassian, pshe

Abassian, zheba

2. Tibetan, bdzhu

Chinese, shi

ADDENDA (1859).

The limited amount of the data must be borne in mind. As has been stated, no vocabularies beyond those of the four works enumerated were used. Had the comparison been more extended, the evidence of the Tibetan affinities of the languages under notice would have been stronger. That this would have been the case has since been proved.

In 1849, just before the publication of my Varieties of Man, I found from my friend Mr. Norris that, upon grammatical grounds, he had come to the same conclusion. A reference to the, then, recently published contributions of Rosen satisfied me that this was the case. The following is an abstract of his exposition of the structure of (1) the Iron, and (2) the Circassian.

(1) IRON.

The Declension of Substantives is as follows;

Singular.Plural.

Nom.fid (father)fid-t`-a

Gen.fid-ifid-t`-i

Dat.fid-énfid-t`-am

Abl.fid-éifid-t`-éi

Nom.moi (husband)moi-t`-a

Gen.moi-imoi-t`-i

Dat.moi-énmoi-t`-am

Abl.moi-éimoi-t`-éi.

The Comparative Degree is formed by the addition of -dar; as chorz=good, chorz-dar=better.

The pronouns of the two first persons are as follows;

1. Az=I. Defective in the oblique cases. Man or ma, defective.

2. Di=Thou. Defective in the nominative singular.

Sing.Plural.

Nom.—mach

Gen.man-imach-i

Dat.man-anmach-én

Accus.manmach

Abl.man-éimach-éi.

Nom.disi-mach

Gen.daw-i[9]si-mach-i

Dat.daw-onsi-mach-én

Accus.dawsi-mach

Abl.daw-éisi-mach-éi.

The signs of the persons of the verbs are -in, -is, -i; -am -ut`, -inc`; e. g.

qus-in=aud-ioqus-am=aud-imus

qus-is=aud-isqus-ut`=aud-itis

qus-i=aud-itqus-inc`=aud-iunt.

The addition of the sound of t helps to form the Irôn preterite. I say helps, because if we compare the form s-ko-t-on=I made, with the root kan, or the form fé-qus-t-on=I heard, with the root qus, we see, at once, that the addition of t is only a part of an inflection.

Beyond this, the tenses become complicated; and that because they are evidently formed by the agglutination of separate words; the so-called imperfect being undoubtedly formed by affixing the preterite form of the word to make. The perfect and future seem to be similarly formed, dele from the auxiliary=be; as may be collected from the following paradigms.

1.

Plural—Present,st-am, st-ut, i-st-i=sumus, estis, sunt.

Singular—Preterite,u-t-an, u-t-as, u-d-i=fui, fuisti, fuit.

Singular—Future,u-gín-an, u-gín-as, u-gén-i=ero, eris, erit.

Imperativefau=esto.

2.

Root, k`an=make. Preterite,=s-k`o-t-on,[10] s-k`o-t-ai, s-k`o-t-a=feci, fecisti, fecit.

3.

Root, kus=hear.

INDICATIVE.

Sing.Plural.

Present,1. qus-inqus-am.

2. qus-isqus-ut`

3. qus-iqus-inc`.

Imperfect,1. qus-ga-k`o-t-onqus-ga-k`o-t-am

2. qus-ga-k`o-t-aiqus-ga-k`o-t-al`

3. qus-ga-k`o-t-aqus-ga-k`o-t-oi

Perfect,1. fé-qus-t-onfé-qus-t-am

2. fé-qus-t-aifé-qus-t-al`

3. fé-qus-t-afé-qus-t-oi

Future,1. bai-qus-g'in-anbai-qus-g'i-stam

2. bai-qus-g'in-asbai-qus-g'i-stut`

3. bai-qus-g'én-ibai-qus-g'i-sti

CONJUNCTIVE.

Sing.Plural.

Present,1. qus-onqus-am

2. qus-aiqus-at`

3. qus-aiqus-oi

Imperfect,1. qus-ga-k`an-onqus-ga-k`an-am

2. qus-ga-k`an-aiqus-ga-k`an-ai`

3. qus-ga-k`an-aqus-ga-k`an-oi

IMPERATIVE.

1. ——bai-qus-am

2. bai-qusbai-qus-ut`

3. bai-qus-abai-qus-oi

Infinitive, qus-in.

Participles, qus-ag, qus-gond, qus-in-ag.

(2) CIRCASSIAN.

In the Absné dialect ab=father, ácĕ=horse; ab ácĕ=father's horse, (verbally, father horse). Here position does the work of an inflection.

The use of prepositions is as limited as that of inflections, sara s-ab ácĕ ist`ap I my-father horse give, or giving am; abna amus`w izbt=wood bear see-did=I saw a bear in the wood; awinĕ wi as`wkĕ=(in) house two doors; ácĕ sis`lit=(on) horse mount I-did.

Hence, declension begins with the formation of the plural number. This consists in the addition of the syllable k`wa.

Acĕ=horse;ácĕ-k`wa=horses.

Atsla=tree;astla-k`wa=trees.

Awinĕ=house;awinĕ-k`wa=houses.

In the pronouns there is as little inflection as in the substantives and adjectives, i. e. there are no forms corresponding to mihi, nobis, &c.

1. When the pronoun signifies possession, it takes an inseparable form, is incorporated with the substantive that agrees with it, and is s- for the first, w- for the second, and i- for the third, person singular. Then for the plural it is h- for the first person, s`- for the second, r- for the third: ab=father;

S-ab=my father;h-ab=our father.

W-ab=thy father;s`-ab=your father.

T-ab=his (her) father;r-ab=their father.

2. When the pronoun is governed by a verb, it is similarly incorporated.

3. Hence, the only inseparable form of the personal pronoun is to be found when it governs the verb. In this case the forms are:

Sa-ra=IHa-ra=we

Wa-ra=thouS`a-ru=ye

Ui=heU-bart`=they.

In sa-ra, wa-ra, ha-ra, s`a-ra, the -ra is non radical. The word u-bart` is a compound.

The ordinal=first is achani. This seems formed from aka=one.

The ordinal=second is agi. This seems unconnected with the word wi-=two; just as in English, second has no etymological connection with two.

The remaining ordinals are formed, by affixing -nto, and (in some case) prefixing -a; as

Cardinals.Ordinals.

3,Chi-ba[11]A-chi-nto

4,P`s`i-baA-p`s`i-nto

5,Chu-baA-chu-nto

6,F-baF-into

7,Bis`-baBs-into

8,Aa-baA-a-nto

9,S`-baS`b-into

10,S`wa-baSw-ento.

In the Absné verbs the distinction of time is the only distinction denoted by any approach to the character of an inflection; and here the change has so thoroughly the appearance of having been effected by the addition of some separate and independent words, that it is doubtful whether any of the following forms can be considered as true inflections.

Root, C'wis`l=ride

1.Present,C'wis`l-ap=I ride[12]=equito.

2.Present,C'wis`l-oit=I am riding.

Imperfect,C'wis`l-an=equitabam.

Perfect,C'wis`l-it=equitavi.

Plusquamperfect,C'wis`l-chén=equitaveram.

Future,C'wis`l-as`t=equitabo.

The person and number is shown by the pronoun. And here must be noticed a complication. The pronoun appears in two forms:—

1st. In full, sara, wara &c.

2nd. As an inseparable prefix; the radical letter being prefixed and incorporated with the verb. It cannot, however, be said that this is a true inflexion.

1.

Sing.1. sara s-c'wisl-oit=I ride

2. wara u-c'wisl-oit=thou ridest

3. ui i-c'wisl-oit=he rides.

2.

Plur.1. hara ha-c'wisl-oit=we ride

2. s`ara s`-c'wisl-oit=ye ride

3. ubart r-c'wisl-oit=they ride

In respect to the name of the class under notice I suggested in 1850 the term Dioscurian from the ancient Dioscurias. There it was that the chief commerce between the Greeks and Romans, and the natives of the Caucasian range took place. According to Pliny, it was carried on by thirty interpreters, so numerous were the languages. The great multiplicity of mutually unintelligible tongues is still one of the characteristics of the parts in question. To have used the word Caucasian would have been correct, but inconvenient. It is already mis-applied in another sense, i. e., for the sake of denoting the so-called Caucasian race, consisting, or said to consist, of Jews, Greeks, Circassians, Scotchmen, ancient Romans, and other heterogeneous elements.

In his paper on the Mongolian Affinities of the Caucasians, published in the Journal of the Asiatic Society of Bengal (1853) Mr. Hodgson has both confirmed and developed the doctrine here indicated—his data on the side of Caucasus being those of the Asia Polyglotta, but those on the side of Tibet and China being vastly augmented; and that, to a great extent, through his own efforts and researches.

Upon the evidence of Mr. Hodgson I lay more than ordinary value; not merely on the strength of his acumen and acquirements in general, but from the fact of his ex-professo studies as a naturalist leading him to over-value rather than under-value those differences of physical conformation that (to take extreme forms) contrast the Georgian and Circassian noble with the Chinese; or Tibetan labourer. Nevertheless, his evidence is decided.

ON THE TUSHI LANGUAGE.

READ BEFORE THE PHILOLOGICAL SOCIETY, FEBRUARY THE 15TH. 1858.

So little light has been thrown upon the languages of Caucasus, that a publication of the year 1856, entitled Versuch über die Thusch-Sprache, by A. Schieffner, may be allowed to stand as a text for a short commentary.

The Tushi is a language belonging to the least known of the five classes into which Klaproth, in his Asia Polyglotta, distributes the languages of Caucasus: viz. (1.) the Georgian. (2.) the Osset or Iron. (3.) the Lesgian. (4.) the Mizhdzhedzhi. And (5.) the Tsherkess or Circassian. It is to the fourth of these that the Tushi belongs; the particular district in which it is spoken being that of Tzowa, where it is in contact with the Georgian of Georgia; from which, as well as from the Russian, it has adopted several words.

The data consist in communications from a native of the district, Georg Ziskorow, with whom the author came in contact at St. Petersburg. They have supplied a grammatical sketch, a short lexicon, and some specimens in the way of composition, consisting of translations of portions of the Gospels, and two short tales of an Arabic or Persian rather than a truly native character. They are accompanied by a German translation.

Taking the groups as we find them in Klaproth, we may ask what amount of illustration each has received in respect to its grammar. In respect to the vocabularies, the Asia Polyglotta gives us specimens of them all.

The Georgian has long been known through the grammar of Maggi, published upwards of two centuries ago. The researches of Rosen on its several dialects are quite recent. Of the Iron there is a copious dictionary by Sjögren, and a short sketch of its grammar by Rosen. The alphabet is Russian, with additions. Rosen has also given a grammatical sketch of the Circassian: This, however, as well as his notice of the Osset, is exceedingly brief. Of the Lesgian we have no grammar at all; and of the Mizhdzhedzhi, or Tshetshent group, the first grammatical sketch is the one before us.

The alphabet is the ordinary Roman modified; the work being addressed to the Russians rather than the natives, and to the European savans in general rather than to the Russians. Otherwise the Georgian alphabet might have been used with advantage; for it is especially stated that the Georgian and Tushi sound-systems are alike. The modifications to which our own alphabet has been subjected, are those that Castrèn has made in his Samoyed grammar and lexicon. So that we may say that it is in Castrèn's Samoyed mode of writing that Schieffner's Tushi grammar and lexicon are exhibited.

In respect to the general relations of the language, the evidence of the work under notice is confirmatory (though not absolutely) of the views to which the present writer has committed himself, viz.—(1.) that the languages of Caucasus in general are so nearly mono-syllabic as to be with fitness designated pauro-syllabic; (2.) that the distinction drawn by Klaproth between the Mizhdzhedzhi and Lesgian groups is untenable; both belonging to the same class, a fact by which the philologic ethnography of Caucasus is, pro tanto, simplified. Upon the first of these points Schieffner writes, that the avoidance of polysyllabic forms has introduced all manner of abbreviations in the language; upon the second, that the little he has seen of the Lesgian grammar induces him to connect it with the Tshetshents. It should be added, however, that in respect to its monosyllabic character, he maintains that the shortness of many of its words is due to a secondary process; so that the older form of the language was more polysyllabic than the present.

Of the chief details, the formation of the cases of the nouns comes first. The declension of the personal pronouns is as follows. With a slight modification it is that of the ordinary substantive as well.

SINGULAR.I.THOU.HE.

Nominativesoḥoo.

Genitivesaiḥaiox̣u.

——————oux̣.

——————ox̣uin.

Dativesonḥonox̣un.

——sona——oux̣na.

Instructiveasaḥox̣us.

——asaaḥaoxuse.

——————oux̣se.

Affectivesoxḥoxox̣ux.

Allativesogoḥogoox̣ugo.

——————oux̣go.

Elativesoxiḥoxioux̣xi.

——————ox̣xi (?).

Comitativesociḥociox̣uci.

——————oux̣ci.

——————ox̣ci (?).

Terminativesogomciḥogomcioux̣gomci.

Adessivesogohḥogoḥoux̣goḥ.

Ablativesogredahḥogredahoux̣gore.

——————oux̣goredah.

PLURAL.WE.YE.THEY.

Nominativewai'txosuobi.

Genitivewai'txaiṡuiox̣ri.

Dativewain'txonṡunox̣arn.

——————suna——

Instructivewaia'txoaiṡox̣ar.

——————aṡiox̣ra.

Affectivewaix'txoxṡuxox̣arx.

Allativewaigo'txogoṡugoox̣argo.

Illativewailo'txoloṡuloox̣arlo.

Elativewaixi'tzoxiṡuxiox̣arxi.

Comitativewaici'txociṡuciox̣arci.

Adessivewaigoh'txogohṡugoḥox̣argoḥ.

Inessive(c.)wailoh'txoloḥṡuloḥox̣arloḥ.

Ablative(c.)waigre'txogreṡugreox̣argore.

————————ox̣ardah.

Elative(c.)wailre'txolreṡulreoḥarlore.

Conversivewaigoih'txogoihṡugoihoḥargoih.

That some of these forms are no true inflexions, but appended prepositions; is speedily stated in the text. If so, it is probable that, in another author or in a different dialect, the number of cases will vary. At any rate, the agglutinate character of the language is indicated. The numerals are—

CARDINAL.ORDINAL.

1.chaduihre.

2.ṡisilǵe.

3.x̣ox̣alǵe.

4.ahewdhewloǵe.

5.ṗxipxilǵe.

6.jeṫxjeixloǵe.

7.worlworloǵe.

8.barlbarloġe.

9.ississloġe.

10.ittittloġe.

11.cha-ittcha-ittloġe.

12.si-ittsi-ittloġe.

19.tqeexçiqeex̣cloġe.

20.tqatqalġe.

This as a word the author connects with the word tqo=also, overagain (auch, wiederum), as if it were 10 doubled, which it most likely is. In like manner tqeexç is one from twenty=undeviginti:—

100 =ṗxauztqa = 5 × 20.

200 =içatatq = 10 × 20.

300 =ṗxiiæatq = 12 × 20.

400 =tqauziq = 20 × 20.

500 =tqauziġ ṗxauztqa = 20 × 20 + 100.

1000 =sac tqauziqa icaiqa = 2 × 400 + 200.

The commonest signs of the plural number are -i and -si, the latter=is in Tshetshents. The suffixes -ne and -bi, the latter of which is found in Lesgian, is stated to be Georgian in origin. No reason, however, against its being native is given.

In verbs, the simplest form is (as usual) the imperative. Add to this -a, and you have the infinitive. The sign of the conditional is ḥe or ḥ; that of the conjunctive ḷe or ḷ.

The tenses are—

(1.) Present, formed by adding -a or -u to the root: i. e. to the imperative form, and changing the vowel.

(2.) Imperfect, by adding -r to the present.

(3.) Aorist, formed by the addition of -r to the

(4.) Perfect; the formation of which is not expressly given, but which is said to differ from the present in not changing the vowel. However, we have the forms xet=find, xeṫi=found; (perf.) xetin=found (aorist). From the participle of the perfect is formed the

(5.) Pluperfect by adding -r.

(6.) The future is either the same as the present, or a modification of it.

I give the names of those moods and tenses as I find them. The language of the Latin grammar has, probably, been too closely imitated.

The first and second persons are formed by appending the pronouns either in the nominative or the instructive form. That an oblique form of the pronoun should appear in the personal inflexion of verbs is no more than what the researches of the late Mr. Garnett, with which we are all so familiar, have taught us to expect. At the same time, the extent to which the instructive and nominative forms are alike must be borne in mind. Let either be appended; and, when so appended, undergo (under certain conditions) certain modifications, and a double origin is simulated. That this is the case in the instances of the work under notice is by no means asserted. The possibility of its being so is suggested.

The participle of the present tense is formed in -in; as dago=eat, dagu-in=eating.

The participle of the preterite ends in -no; as xac̣e=hear, xac̣-no=heard.

There are auxiliary verbs, and no small amount of euphonic changes; of which one, more especially, deserves notice. It is connected with the gender of nouns. When certain words (adjectives or the so-called verb substantive) follow certain substantives, they change their initial. Thus ḥaṫxleen wa=the prophet is, ḥaṫxleensi ba=the prophets are, waṡo wa=the brother is, waṡar ba=the brothers are.

Again—naw ja=the ship is, nawr ja=the ships are; bstiuno ja=the wife is, bstee da=the wives are.

This is said to indicate gender, but how do we know what gender is? The words themselves have neither form nor inflexion which indicates it. Say that instead of gender it means sex, i. e. that the changes in question are regulated by natural rather than grammatical characters. We still find that the word naw is considered feminine—feminine and inanimate. This, however, is grammatical rather than natural, sex—"das weibliche Geschlecht wird bey unbelebten Gegenständen auch im Plural durch j-, bei belebten durch a ausgedrückt." Then follow the examples just given. How, however, do we know that these words are feminine? It is submitted that the explanation of this very interesting initial change has yet to be given. It recalls, however, to our memory the practice of more languages than one, the Keltic, the Woloff, the Kafre, and several other African tongues, wherein the change is initial, though not always on the same principle.

So, also, the division of objects into animate and inanimate recalls to our mind some African, and numerous American, tongues.

Such is the notice of the first of the Mizhdzhedzhi or Tshetshents (we may say Lesgian) forms of speech of which the grammatical structure has been investigated; a notice which suggests the question concerning its affinities and classification.

The declension points to the Ugrian, or Fin, class of languages; with which not only the Tshetshents, but all the other languages of Caucasus have long been known to have miscellaneous affinities. The resemblance, however, may be more apparent than real. The so-called cases may be combinations of substantives and prepositions rather than true inflexions, and the terminology may be more Ugrian in form than in reality. Even if the powers of the cases be the same, it will not prove much. Two languages expressing a given number of the relations that two nouns may bear to each other will, generally, express the same. Cases are genitive, dative and the like all the world over—and that independent of any philological affinity between the languages in which they occur. The extent to which they are also Caritive, Adessive and the like has yet to be investigated.

The Ugrian affinities, then, of the Tshetshents are indirect; it being the languages of its immediate neighbourhood with which it is more immediately connected. In the way of vocabularies the lists of the Asia Polyglotta have long been competent to show this. In the way of grammar the evidence is, still, far from complete. The Georgian, to which Maggi gives no more than six cases, has a far scantier declension than the Tushi, at least as it appears here. The Circassian, according to Rosen, is still poorer.

In the verbs the general likeness is greater.

In the pronouns, however, the most definite similarity is to be found; as may be seen from the following forms in the Circassian:—

Ab=father.

1.S-ab=my father.2.H-ab=our father.

W-ab=thy father. S'-ab=your father.

L-ab=his father. S-ab=their father.

To which add—

Sa-ra=I.Ha-ra=we.

Wa-ra=thou.S'a-ra=ye.

Ui=he.U-bart=they.

The amount of likeness here is considerable. Over and above the use of s for the first person singular, the s' in the second person plural should be noticed. So should the b and r in the Circassian u-bart; both of which are plural elements in the Tushi also.

Finally (as a point of general philology), the double forms of the Tushi plurals wai and tx̣o suggest the likelihood of their being exclusive and inclusive; one denoting the speaker but not the person spoken to, the other both the person spoken to and the person who speaks; plurals of this kind being well known to be common in many of the ruder languages.

ON THE NAME AND NATION OF THE DACIAN KING DECEBALUS, WITH NOTICES OF THE AGATHYRSI AND ALANI.

READ BEFORE THE PHILOLOGICAL SOCIETY, APRIL 17TH 1854.

The text of Herodotus places the Agathyrsi in Transylvania (there or thereabouts). (See F. W. Newman On Scythia and the surrounding Countries, according to Herodotus, Philological Society's Proceedings, vol. i. p. 77.)

The subsequent authors speak of them as a people who painted (tattooed?) their bodies; the usual epithet being picti.

The same epithet is applied to the Geloni; also a population of the Scythia of Herodotus.

For accurate knowledge the locality of the Agathyrsans was too remote—too remote until, at least, the date of the Dacian wars; but the Dacian wars are, themselves, eminently imperfect in their details, and unsatisfactory in respect to the authorities for them.

There is every reason, then, for a nation in the locality of the Agathyrsi remaining obscure—in the same predicament (say) with the Hyperborei, or with the occupants of Thule.

But there is no reason for supposing the obliteration of the people so called; nor yet for supposing a loss of its name, whether native or otherwise.

Hence, when we get the details of Dacia we may reasonably look out for Agathyrsi.

How far must we expect to find their name unmodified? This depends upon the population through whom the classical writers, whether Latin or Greek, derived it. Now it is submitted, that if we find a notice of them in the fifth century A. D., and that in an account relating to Dacia and Pannonia, the medium has, probably, been different from that through which Herodotus, amongst the Greek colonies of the Black Sea, obtained his accounts. The details of this difference of medium are not very important, and the discussion of them would be episodical to the present paper, if not irrelevant. It is enough to remark, that a difference of medium is probable; and, as a consequence thereof, a difference in the form of the name.

This is preliminary and introductory to the notice of the following passage of Priscus, to whom we owe the account of one of the embassies to Attila—Ὁ πρεσβύτερος ἧρχε τῶν Ἀκατζίρων καὶ τῶν λοιπῶν ἐθνῶν νεμομένων τὴν πρὸς τὸν Πόντον Σκυθικέν. Another form (also in Priscus) is Ἀκατίροι. They are specially called Akatiri Hunni. Jornandes' form is Acatziri.

Place for place, this gives us the Agathyrsi of Herodotus as near as can be expected; and, name for name it does the same: the inference being that the Akatziri of Priscus are the descendants of the Agathyrsi of Herodotus. Of course, evidence of any kind to the migration, extinction, or change of name on the part of the population in question would invalidate this view. Such evidence, however, has not been produced, nor has the present writer succeeded in finding, though he has sought for it.

Descendants then of the Agathyrsi, and ancestors of the Akatziri may have formed part of the population of Dacia when Domitian and Trajan fought against Decebalus; a part that may have been large or small, weak or powerful, homogeneous with the rest of Dacia or different from it. Assuming it to have been different, it may still have supplied soldiers—even leaders. Decebalus himself may as easily have belonged to the Agathyrsan part of Dacia as to any other. A very little evidence will turn the balance in so obscure a point as the present.

Now, no German and no Slavonic dialects give us either the meaning of the name Decebalus or any name like it. It stands alone in European history. Where does it appear? In the history of the Turks. The first known king of the Turks bears the same name as the last of the Dacians. Dizabulus (Διζαβούλος) was that khan of the Turks of Tartary to whom Justinian sent an embassy when the Avars invaded the Eastern empire.

This (as is freely admitted) is a small fact, if taken alone; but this should not be done. The cumulative character of the evidence in all matters of this kind should be borne in mind, and the value of small facts measured by the extent to which they stand alone, or are strengthened by the coincidence of others. In the latter case they assume importance in proportion to the mutual support they give each other; the value of any two being always more than double that of either taken singly.

On the other hand, each must rest on some separate substantive evidence of its own. To say that Decebalus was an Agathyrsan because the Agathyrsans were Turks, and that the Agathyrsans were Turks because Decebalus was one of them, is illegitimate. There must be some special evidence in each case, little or much.

Now the evidence that the Agathyrsi were Turks lies in the extent to which (a) they were Scythians (Skoloti), and (b) the Scythians (Skoloti) were Turks;—neither of which facts is either universally admitted or universally denied. The present writer, however, holds the Turk character of the Agathyrsi on grounds wholly independent of anything in the present paper; indeed, the suggestion that the Acatziri are Agathyrsi is, not his, but Zeuss'.—(See Die Deutschen and die Nachbarstämme, v. Bulgari, p. 714.)

If Agathyrs- be Akatzir- in some older, what is the latter word in any newer form?—for such there probably is. Word for word, it is probably the same as Khazar, a denomination for an undoubtedly Turk tribe which occurs for the first time in Theophanes:—Τοῦρκοι ἀπὸ τῆς ἐώας οὓς Χαζάρους ὀνομάζουσιν. This is A. D. 626. Whether, however, the same populations were denoted is uncertain. There are certain difficulties in the supposition that they were absolutely identical.

It is not, however, necessary that they should be so. There might be more than one division of a great stock, like the Turk so called. Nay, they might have been populations other than Turk so designated, provided only that there were some Turk population in their neighbourhood so to call them. More than this. The word may be current at the present moment, though, of course, in a modified form. Suppose it to have been the Turk translation of pictus; or rather, suppose the word pictus to be the Latin translation of Agathyrs- (Akatzir-): what would the probable consequence be? Even this, that wherever there was a painted (or tattooed) population in the neighbourhood of any member of the great Turk stock, the name, or something like it, might arise. Be it so. If the members of the same Turk stock lay wide apart, the corresponding painted or tattooed populations lying wide apart also might take the same name.

The details suggested by this line of criticism may form the subject of another paper. In the present, the author hazards a fresh observation—an observation on a population often associated with the Agathyrsi, viz. the Geloni. Seeing that we have such forms as Unni (the Greek form is Οὖννοι, not Οὗννοι) and Chuni (=Huns); Arpi and Carpi; Attuarii and Chattuari, &c.; and seeing the affinity between the sounds of g and k; he believes that the word Geloni may take another form and begin with a vowel (Elôni, Alôni). Seeing that their locality is nearly that of the Alani of a latter period; seeing that the middle syllable in Alani (in one writer at least) is long—ἀλκήεντες Ἀλαῦνοι; seeing that Herodotus, who mentions the Geloni, knows no Alani, whereas the authors who describe the Alani make (with one exception about to be noticed) no mention of the Geloni, he identifies the two populations, Geloni and Alani, or vice versâ. He deduces something more from this root l—n (λ—ν). Let the name for the Alans have reached the Greeks of the Euxine through two different dialects of some interjacent language; let the form it took in Greek have been parisyllabic in one case, whereas it was imparisyllabic in the other, and we have two plurals, one in -οι, as Γέλωνοι, Ἄλαυνοι, Ἄλανοι, and another in -ες, as Γέλωνες, Ἄλαυνες, Ἄλανες,—possible, and even probable, modifications of the original name, whatever that was. Now, name for name, Αλανες comes very near Ελληνες; and in this similarity may lie the explanation of the statement of Herodotus as to the existence of certain Scythian Greeks (ἑλληνες Σκυθαι)—iv. 17. 108.

If so these Scythian Greeks were Alans.

The exception, indicated a few lines above, to the fact of only one author mentioning both Geloni and Alani, is to be found in Ammianus Marcellinus (xxxi. 2. 13. 14). The passage is too long to quote. It is clear, however, that whilst his Alani are spoken of from his own knowledge, his Geloni are brought in from his book-learning, i. e. from Herodotus.

NOTES.

Note 1.

Evidence of any kind to the migration, extinction or change of name on the part of the populations in question would invalidate this view. Such evidence has not been produced &c.—The fuller consideration of the question involved in this statement is to be found in Dr. W. Smith's Dictionary of Greek and Roman Geography vv. Hunni, Scythia, and Sarmatia.

Note 2.

The details suggested by this line of criticism &c.—These are to the effect that in the word Agathyrsi we get an early Turk gloss, of which the history is somewhat curious. It exists, at the present moment in England, having come via Hungary. It exists in Siberia, on the very frontier of the America.

It is the English word Hussar=Khazar. Here we have it in its abbreviated form.

It is the Siberian word Yukahir, Yukazhir, or Yukadzhir.

The "native name of the Yukahiri of Siberia is Andon Domni. The Koriaks call them Atal. Their other neighbours are the Turk Yakuts. Hence it is probable that it is to the Yakut language that the term Yukahir (also Yukadzhir) is referrible. If so, its probable meaning is the same as the Koriak Atal, which means spotted. It applies to the Yukahiri from their spotted deerskin dresses.

Now, south of these same Yakuts, who are supposed to call the Andon Domni by the name Yukahiri (or Yukadzhiri), live a tribe of Tungusians. These are called Tshapodzhir—but not by themselves. By whom? By no one so probably as by the Yakuts. Why? Because they tattoo themselves. If so, it is probable that Yukadzhir and Tshapodzhir are one and the same word; at any rate, a likely meaning in a likely language has been claimed for it.

Let it, then, be considered as a Turk word, meaning spotted, tattooed, painted,—provisionally. It may appear in any part of the Turk area, provided only, that some nation to which one of the three preceding adjectives applies be found in its neighbourhood. It may appear, too, in any state of any Turk form of speech. But there are Turk forms of speech as far distant from the Lena and Tunguska as Syria or Constantinople; and there are Turk glosses as old as Herodotus. One of these the present writer believes to be the word Agathyrsi, being provided with special evidence to shew that the nation so called were either themselves Turks or on a Turk frontier. Now, the Agathyrsi are called the picti Agathyrsi; and it is submitted to the reader that the one term is the translation of the other—the words Agathyrs (also Akatzir), Yukadzhir, and Tshapodzhir, being one and the same."—From the author's Native Races of the Russian Empire.

ON THE LANGUAGE OF LANCASHIRE, UNDER THE ROMANS.

READ BEFORE THE HISTORICAL SOCIETY OF LANCASHIRE AND CHESHIRE. 8TH JANUARY, 1857.

In the present paper, advantage is taken of the local character of the Historic Society of Lancashire and Cheshire, to make the name of the county serve as a special text for a general subject. What applies to Lancashire applies to any county in Roman England.

The doctrine is as follows—that in Lancashire particularly, and in England in general, the predominant language for the first five centuries of our era was not Latin but British.

The writer is so far from laying this down as a novelty, that he is by no means certain, that it may not be almost a truism. He is by no means certain, that there is a single one of those to whom he addresses himself, who may now hold, or even have held, the opposite opinion. He is fully aware that excellent authorities have maintained both sides of the question. He is only doubtful as to the extent to which the one doctrine may preponderate over the other.

If the question were to be settled by an appeal to the history of the more influential opinions concerning it, we should find that, in a reference to the earliest and the latest of our recent investigators, Dr. Prichard would maintain one side of the question, Mr. Wright another. The paper of the latter, having been printed in the Transactions of the Society, is only alluded to. The opinion of Dr. Prichard is conveyed in the following extract—"The use of languages really cognate must be allowed to furnish a proof, or at least a strong presumption, of kindred race. Exceptions may indeed, under very peculiar circumstances, occur to the inference founded on this ground. For example, the French language is likely to be the permanent idiom of the negro people of St. Domingo, though the latter are principally of African descent. Slaves imported from various districts in Africa, having no common idiom, have adopted that of their masters. But conquest, or even captivity, under different circumstances, has scarcely ever exterminated the native idiom of any people, unless after many ages of subjection; and even then, vestiges have perhaps always remained of its existence. In Britain, the native idiom was nowhere superseded by the Roman, though the island was held in subjection upwards of three centuries. In Spain and in Gaul, several centuries of Latin domination, and fifteen under German and other modern dynasties, have proved insufficient entirely to obliterate the ancient dialects, which were spoken by the native people before the Roman conquest. Even the Gypsies, who have wandered in small companies over Europe for some ages, still preserve their original language in a form that can be everywhere recognised."[13]

Upon the whole, I think that the current opinion is in favour of the language of Roman Britain having been Latin; at any rate I am sure that, before I went very closely into the subject, my own views were, at least, in that direction. "What the present language of England would have been, had the Norman conquest never taken place, the analogy of Holland, Denmark, and many other countries enables us to determine. It would have been as it is at present. What it would have been had the Saxon conquest never taken place, is a question wherein there is far more speculation. Of France, of Italy, of Wallachia, and of the Spanish Peninsula, the analogies all point the same way. They indicate that the original Celtic would have been superseded by the Latin of the Conquerors, and consequently that our language, in its later stages, would have been neither British nor Gaelic, but Roman. Upon these analogies, however, we may refine. Italy was from the beginning, Roman; the Spanish Peninsula was invaded full early; no ocean divided Gaul from Rome; and the war against the ancestors of the Wallachians was a war of extermination."[14]

In these preliminary remarks we find a sufficient reason for going specially into the question; not, however, as discoverers of any new truth, nor as those who would correct some general error, but rather, in a judicial frame of mind, and with the intention of asking, first, how far the actual evidence is (either way) conclusive; next, which way (supposing it to be inconclusive) the presumption lies; and thirdly, what follows in the way of inference from each of the opposing views.

What are the statements of the classical writers, subsequent to the reduction of Britain, to the effect that the Romans, when they conquered a Province, established their language? I know of none. I know of none, indeed, anterior to the Britannic conquest. I insert, however, the limitation, because in case such exist, it is necessary to remember that they would not be conclusive. The practice may have changed in the interval.

Is there anything approaching such a statement? There is a passage in Seneca to the effect "that where the Roman conquers there he settles."

But he conquered Britain. Therefore he established his language. Add to this that where he established his own language, there the native tongue became obliterated. Therefore the British died off.

If so, the Angles—when they effected their conquest—must have displaced, by their own English, a Latin rather than a British, form of speech.

But is this the legitimate inference from the passage in question? No. On the contrary, it is a conclusion by no means warranted by the premises. Nevertheless, as far as external testimony is concerned, there are no better premises to be found.

But there is another element in our reasoning. In four large districts at least,—in the Spanish Peninsula, in France, in the Grisons, and in the Danubian Principalities—the present language is a derivative from the Latin, which was, undoubtedly and undeniably, introduced by the Roman conquest. From such clear and known instances, the reasoning to the obscure and unknown is a legitimate analogy, and the inference is that Britain was what Gallia, Rhætia, Hispania, and Dacia were.

In this we have a second reason for the fact that there are many who, with Arnold, hold, that except in the particular case of Greece, the Roman world, in general, at the date of the break-up of the Empire, was Latin in respect to its language. At any rate, Britannia is reasonably supposed to be in the same category with Dacia—a country conquered later.

On the other hand, however, there are the following considerations.

I. In the first place the Angle conquest was gradual; so gradual as to give us an insight into the character of the population that was conquered. Was this (in language) Latin? There is no evidence of its having been so. But is there evidence of its having been British? A little. How much, will be considered in the sequel.

II. In the next place the Angle conquest was (and is) incomplete; inasmuch as certain remains of the earlier and non-Angle population still exist. Are these Latin? Decidedly not; but on the contrary British,—witness the present Britons of Wales, and the all but British Cornish-men, who are now British in blood, and until the last century were, more or less, British in language as well.

But this is not all. There was a third district which was slow to become Angle, viz.: part of the mountain district of Cumberland and Westmoreland. What was this before it was Angle? Not Roman but British.

Again—there was a time when Monmouthshire, with (no doubt) some portion of the adjoining counties, was in the same category in respect to its non-Angle character with Wales. What was it in respect to language? Not Roman but British.

Again—mutatis mutandis. Devonshire was to Cornwall as Monmouth to Wales. Was it Roman? No—but, on the contrary, British.

Now say, for the sake of argument, that Cornwall, Wales, and Cumberland were never Roman at all, and consequently, that they prove nothing in the question as to the introduction of the Latin language. But can we say, for even the sake of argument, that Devon and Monmouth were never Roman? Was not, on the contrary, Devon at least, exceedingly Roman, as is shewn by the importance of Isca Danmoniorum, or Exeter.

Or, say that the present population of Wales is no representative of the ancient occupants of that part of Britain, but, on the contrary, descended from certain immigrants from the more eastern and less mountainous parts of England. I do not hold this doctrine. Admitting it, however, for the sake of argument—whence came the present Welsh, if it came not from a part of England where British, rather than Latin, was spoken? There must have been British somewhere; and probably British to the exclusion of Latin.

The story of St. Guthlac of Croyland is well-known. It runs to the effect that being disturbed, one night, by a horrid howling, he was seriously alarmed, thinking that the howlers might be Britons. Upon looking-out, however, he discovered that they were only devils—whereby he was comforted, the Briton being the worse of the two. Now the later we make this apocryphal story, the more it tells in favor of there having been Britons in Lincolnshire, long after the Angle conquest. Yet Lincolnshire (except so far as it was Dane,) must have been one of the most Angle portions of England. In France, Spain, Portugal, the Grisons, Wallachia or Moldavia, such devils as those of St. Guthlac would have been Romans.

As the argument, then, stands at present, we have traces of the British as opposed to the Angle, but no traces of the Latin in similar opposition.

Let us now look at the analogies, viz: Spain, (including Portugal,) France, Switzerland and the Danubian Principalities; in all of which we have had an aboriginal population and a Roman conquest, in all of which, too, we have had a third conquest subsequent to that by Rome—even as in Britain we have had the triple series of (A) native Britains, (B) Roman conquerors, (C) Angles.

What do we find? In all but Switzerland, remains of the original tongue; in all, without exception, remains of the language of the population that conquered the Romans; in all, without exception, something Roman.

In Britain we find nothing Roman; but, on the contrary, only the original tongue and the language of the third population.

I submit that this is strong primâ facie evidence in favour of the Latin having never been the general language of Britain. If it were so, the area of the Angle conquest must have exactly coincided with the area of the Latin language. Is this probable? I admit that it is anything but highly improbable. The same practicable character of the English parts of Britain (as opposed to the Welsh, Cornish, and Cumbrian) which made the conquest of a certain portion of the Island easy to the Romans as against the Britons, may have made it easy for the Angles as against the Romans; and vice versa, the impracticable character of Wales, Cornwall, and Cumberland, that protected the Britons against their first invaders, may have done the same for them against the second. If so, the two areas of foreign conquest would coincide. I by no means undervalue this argument.

It is almost unnecessary to say that the exact conditions under which Britain was reduced were not those of any other Roman Province.

In respect to Spain, the Roman occupancy was early, having begun long before that of Northern and Central Gaul, having begun during the Punic wars, and having become sufficiently settled by the time of Augustus to command the attention of Strabo on the strength of the civilization it had developed. In Spain, then, there was priority in point of time to account for any extraordinary amount of Roman influences.

Gaul, with the exception of the earlier acquisitions in the Narbonensis, was the conquest of one of the most thorough-going of conquerors. The number of enemies that Cæsar slaughtered has been put at 1,000,000. Without knowing the grounds of this calculation, we may safely say that his campaigns were eminently of a destructive character.

The conquerors of the Breuni, Genauni, and similar occupants of those parts of Switzerland where the Rumonsch Language (of Latin origin) is now spoken, were men of similar energy. Neither Drusus nor Tiberius spared an enemy who opposed. Both were men who would "make a solitude and call it peace."

That Trajan's conquest of Dacia was of a similar radical and thorough-going character is nearly certain.

Now, the evidence that the conquests of the remaining provinces were like those of the provinces just noted, is by no means strong. At the same time, it must be admitted that the analogy established by four such countries as Gaul, Spain, Switzerland, and Moldo-Wallachia is cogent. What was the extent to which Africa, Pannonia, Illyricum, Thrace, and the Mœsias were Romanized? Of Asia? I say nothing. It was sufficiently Greek to have been in the same category with Greece itself, and in Greece itself we know that no attempts were made upon the language.

Africa was Latin in its literature; and, at a later period, pre-eminently Latin in its Christianity. But the evidence that the vernacular language was Latin is nil, and the presumptions unfavourable. The Berber tongue of the present native tribes of the whole district between Egypt and the Atlantic is certainly of high antiquity; it being a well-known fact, that in it, several of the names in the geography of classical Africa are significant. Now this is spread over the country indifferently. Neither does it show any notable signs of Latin intermixture. Neither is there trace, or shadow of trace, of any form of speech of Latin origin throughout the whole of Tunis, Tripoli, Algiers or Morocco.

In Pannonia and Illyricum, the same absence of any language of Latin origin is manifest. Pannonia and Illyricum have had more than an average amount of subsequent conquerors and occupants—Goths, Huns, Avars, Bulgarians, Slavonians, Hungarians, Germans. That the Slovak, however, in the north, and the Dalmatian forms of the Servian in the south, represent the native languages is generally admitted—now, if not long ago. These, then, have survived. Why not, then, the Latin if it ever took root?

In respect to Thrace, it is just possible that it may have been, in its towns at least, sufficiently Greek to have been in the same category with Greece proper. I say that this is just possible. In reality, however, it was more likely to be contrasted with Greece than to be classed with it. One thing, however, is certain, viz.:—that the country district round Constantinople was never a district in which Latin was vernacular. Had it been so, the fact could hardly have been unnoticed, or without influence on the unequivocally Greek Metropolis of the Eastern Empire.

If the doctrine that Thrace may have been sufficiently Greek to forbid the introduction of the Latin be doubtful, the notion that the Mœsias were so is untenable. Yet the Latin never seems to have been vernacular in either of them. Had it been so, it would probably have held its ground, especially in the impracticable mountains and forests of Upper Mœsia or the modern Servia. Yet where is there a trace of it? Of all the Roman Provinces, Servia or Upper Mœsia seems to be the one wherein the evidence of a displacement of the native, and a development of a Latin form of speech, is at its minimum, and the instance of Servia is the one upon which the analogous case of Britain best rests.

The insufficiency of the current reasons in favour of the modern Servian being of recent introduction have been considered by me elsewhere.

Now comes the notice of a text which always commands the attention of the ethnological philologue, when he is engaged upon the Angle period of our island's history. It refers to the middle of the eighth century, the era of the Venerable Beda, from whose writings it is taken. I give it in extenso. It runs "Hæc in presenti, juxta numerum librorum quibus lex divina scripta est, quinque gentium linguis, unam eandemque summæ veritatis et veræ sublimitatis scientiam scrutatur et confitetur; Anglorum, videlicet, Brittonum, Scottorum, Pictorum et Latinorum quæ meditatione scripturarum, cæteris omnibus est facta communis".[15]

That the Latin here is the Latin of Ecclesiastical, rather than Imperial, Rome, the Latin of the Scriptures rather than classical writers, the Latin of a written book rather than a Lingua Rustica, is implied by the context.

Should this, however, be doubted, the following passage, which makes the languages of Britain only four, is conclusive—"Omnes nationes et provincias Brittanniæ, quæ in quatuor linguas, id est Brittonum, Pictorum, Scottorum et Anglorum divisæ sunt, in ditione accepit."[16]

It is the first of these two statements of Beda's that the following extract from Wintoun is founded on.

Cronykil, I. xiii, 39.

But the Latin of the scriptures may have been the Latin of common life as well. Scarcely. The change from the written to the spoken language was too great for this. What the latter would have been we can infer. It would have been something like the following "Pro Deo amur et pro Xristian poblo et nostro commun salvament d'ist di en avant, in quant Deus savir et poder me dunat, si salvarai eo cist meon fradre Karlo, et in ajudha et in cadhuna cosa, si com om per dreit son fradre salvar dist, in o quid il me altresi fazet: et ab Ludher nul plaid nunquam prindrai uni, meon vol, cist meon fradre Karle in damno sit."

This is the oath of the Emperors Karl and Ludwig, sons of Charlemagne, as it was sworn by the former in A. D. 842. It is later in date than the time of Beda by about a century; being in the Lingua Rustica of France. Nevertheless, it is a fair specimen of the difference between the spoken languages of the countries that had once been Roman Provinces and the written Latin. Indeed, it was not Latin, but Romance; and, in like manner, any vernacular form of speech, used in Britain but of Roman origin, would have been Romance also.

The conclusion which the present notice suggests is—

That the testimony of authors tells neither way.

That the presumptions in favour of the Latin which are raised by the cases of Gaul, Spain, Rhætia, and Dacia, are anything but conclusive.

That the inferences from the earliest as well as the latest data as to the condition of English Britain, the inferences from the Angle conquest, and the inferences from the present language of Wales, are decidedly against the Latin.

I may, perhaps, be allowed to conclude by a reference to a paper already alluded to, as having been laid before the present Society, by Mr. Wright. This is to the effect, that the Latin reigned paramount not only in England, but in Wales also, under the Roman dominion; the present Welsh being of recent introduction from Armorica.

That the population was heterogeneous is certain, the Roman Legionaries being, to a great extent, other than Roman. It is also certain that there was, within the island, at an early period, no inconsiderable amount of Teutonic blood. It is certain, too, that the name Briton had different applications at different times.

If so, the difference between Mr. Wright and myself, in respect to the homogeneousness or heterogeneousness of the Britannic population, is only a matter of degree.

In respect to the particular fact, as to whether the British or Latin language was the vernacular form of speech, we differ more decidedly. That the British was unwritten and uncultivated is true; so that the exclusive use of the Latin for inscriptions is only what we expect. The negative fact that no British name has been found inscribed, I by no means undervalue.

The preponderance, however, of a Non-British population, and the use of the Latin as the vernacular language, are doctrines, which the few undoubted facts of our early history impugn rather than verify.

The main difficulty which Mr. Wright's hypothesis meets—and it does meet it—lies in the fact of the similarity between the Welsh and Armorican being too great for anything but a comparatively recent separation to account for. Nevertheless, even this portion of what may be called the Armorican hypothesis, is by no means incompatible with the doctrine of the present paper. The Celtic of Armorica may as easily have displaced the older Celtic of Britain (from which, by hypothesis, it notably differed) as it is supposed to have displaced the Latin.

I do not imagine this to have been the case; indeed I can see reasons against it, arising out of the application of Mr. Wright's own line of criticism.

I think it by no means unlikely that the argument which gives us the annihilation of the British of the British Isles, may also give us that of the Gallic of Gaul. Why should Armorica have been more Celtic than Wales? Yet, if it were not so, whence came the Armorican of Wales? I throw out these objections for the sake of stimulating criticism, rather than with the view of settling a by no means easy question.