Sept. 2, 1880] 



NATURE 



421 



are the ethnical relations of the typical n.au of South 

 Wales? 



Nine people out of every ten to whom this question might be 

 addressed would unhesitatingly answer that the true Welsh are 

 Celts or Kelts.' And they would seek to justify their answer by 

 a confident appeal to the Welsh language. No one has any 

 doubt about the position of this language as a member of the 

 Keltic family. The Welsh and the Breton fall naturally together 

 as living members of a group of languages to which Prof. Rhys 

 applies the term Brythoiiic^ a group which also includes such 

 fossil tongues as the old Cornish, the speech of the Strathclyde 

 Britons, and possibly the language of the Picts and of the Gauls. 

 On the other hand, the Gaelic of Scotland, the Irish, and the 

 Manx arrange themselves as naturally in another group, which 

 Prof. Rhys distinguishes as the Coidelic branch of the Keltic 

 stock." But does it necessarily follow that all the people; who 

 are cloFcly linked together by speaking, or by having at some 

 time spoken, these Keltic languages, are as closely linked to- 

 gether by ties of blood ? Great as the value of language unques- 

 tionably is as an aid to ethnological classification, are we quite 

 safe in concluding that all the Keltic-speaking peoples are one in 

 race ? 



The answer to such a question must needs depend upon the 

 sense in which the anthropologist uses the word Kelt. History 

 and tradition, philology and ethnology, archiEoIogy and cranio- 

 logy, have at different times given widely divergent definitions of 

 the term. Sometimes the word has been used with such elasticity 

 as to cover a multitude of peoples who differ so widely one from 

 another in physical characteristics that if the hereditary persist- 

 ence of such qualities counts for anything, tliey cannot possibly 

 he referred to a common stock. Sometimes, on the other hand, 

 the word has been so restricted in its definition that it has actually 

 e.\cluded the most typical of all Kelts — the Gaulish Kelts of 

 Caesar. According to one authority the Kelt is short, according 

 to another tall ; one ethnologist defines him as being dark, 

 another as fair ; this craniologist finds that he has a long skull, 

 while that one declares that his skull is short. It was no doubt 

 this ambiguity that led so keen an observer as Dr. Eeddoe to 

 remark, nearly fifteen years ago, that "Kelt and Keltic are 

 terms which were useful in their day, but which have ceased to 

 convey a distinct idea to the mir.ds of modern students." ^ 



No anthropologist has laboured more persistently in endea- 

 vouring to evoke order out of this Keltic chaos than the late Paul 

 Broca. What, let us ask, was the opinion of this distinguished 

 anthropologist on the Keltic question?^ Prof. Broca always held 

 that the name of Kelt should be. 'trictly limited to the Kelt of posi- 

 tive history — to the people, or rather confederation of peoples, 

 actually seen by Cssar in Keltic Gaul— and, of course, to their 

 descendants in the same area. Every schoolboy is familiar with 

 'he epitome of Gaulii-h ethnology given by Julius in his opening 

 chapter. Nothing can be clearer than his description of the 

 tripartite division of Gaul, and of the separation between the th'-ee 

 peoples who inhabited the country — the Belgre, the Aquitani, and 

 the Celt.Te. Of these three peoples the most important were those 

 whom the Romans called Calli, but w ho called themselves, as 

 the historian tells us, Cclta. The countiy occupied by the Keltic 

 population stretched from the Alps to the Atlantic in one direc- 

 tion, and from the Seine to the Garonne in another ; but it is 

 difficult to find any direct evidence that the Kelts of this area ever 

 crossed into Britain. Broca refused to apply the name of Kelt 

 to the old inhabitants of Belgic Gaul, and as a matter of course 



^ \yhether this word should be written Celt or Kelt seems to be a matter 

 of scientific indifference. Probably the balance of opinion among ethno- 

 logists^ is in the direction of the former rendering. Nevertheless it must be 

 borne in mind that the word "celt" is so commonly used nowadays by 

 writers on prehistoric anthropology to designate an axe-head, or some such 

 weapon, whether of meta! or of stone, that it is obviously desirable to make 

 the difference between the archaeological word and the ethnological term as 

 clear as possible. If ethnologists persist in writing " Celt," the two words 

 differ only in the magnitude tf an initial, and when spoken are absolutely 

 indjstinguishable. 1 shall therefore write, as a matter of expediency, 

 ■' Kelt." It may be true, as Mr Knight Watson has pointed out, that 

 there was originally no justification for using the word "celt" as the name 

 cf a weapon, but it is too late in the day to attempt to oust so deeply-rooted 

 a word from the vocabul.-xry of the archaologists. 



' "Lectures on Welsh Philology," by John Rhys, M.A., 2nd edition, 

 1878, p. IS. 



5 Mem. Anthrop. Sec. Lon., vol. ii. 1E66, p. 348. 



4 The following are Broca's principal contributions to this vexed ques- 

 tion :— " Qu'est-ce que les Celtes? " Btillftitis dcia SocUte Anthropologioue 

 de Paris, t. v. p. 457 ; " Le Norn des Celtes," Hid. 2 ser. t. ix. p. 662 ; " Sur 

 les Textes relatifs aux Celtes dans le Grande-Bretagne," ibid. 2 sc'r. t. xii. 

 p. 509: " La Race Celtique, ancienne et moderne," Rez'uc d^ Anihrfpologiet 

 t. ii. p. 57S ; and " Recherches sur I'Ethnologie de la France," Mhn. de la 

 See. Anthrop., t. i. p. i. 



he denied it to any of the inhabitants of the British Isles. 

 Writing as late as 1877, in full view of all the arguments which 

 had been adduced .against his opinions, he still said: "Je con- 

 tinue a soutenir, ju^qu'a preuve du contraire, ce que j'ai avance il 

 y adouze, ans dans notre premiere discussion sur les Celtes, savoir, 

 qu'il n'existe aucune preuve, qu'on ait constate dans les Iles- 

 Britanniques I'existence d'un peuple portant le nom de Celtes."' 



Nevertheless, in discussing the Keltic question with M. Henri 

 Martin, he admitted the convenience, almost the propriety, of 

 referring to all who spoke Keltic languages as Keltic peoples, 

 though of course he would not hear of their being called Kelts. 

 " On pent tres bien les nommer les peuples celtiques. Mais il 

 est entierement faux de les appeler les Cdles, comme on le fait 

 si souvent."- As to the eminent historian himself, I need hardly 

 say that M. Martin adheres to the popular use of the word Kelt, 

 and even goes so far as to speak of the county in which we are 

 now assembled as "le Glamorgan, le pays aujourd'hui le plus 

 celtique de I'Europe." ^ 



Whether we use the word Kelt in its wide linguistic sense or 

 in the narrower sense to which it has been reduced by the French 

 anthropologists, it is important to remember that the Welsh do 

 not designate, and never have designated themselves by this 

 term or by any similar word. Their national name is Cymry, 

 the plural of Cymro. My former colleague, the Rev. Prof. 

 Silvan Evans, kindly informs me that the most probable deriva- 

 tion of this word is from cyd- (the d being changed to m for 

 assimilation with the following b, like the n of its Latin cognate 

 con) and Inv-, "country," the old form of which is brog, as 

 found in AUobrogir, and some other ancient names. The 

 meaning of Cyritry is therefore "fellow-countrymen," or com- 

 patriots. Such a meaning naturally suggests that the name 

 must have been assumed in consequence of some foreign in- 

 vasion — possibly when the WeUh were banded together against 

 either the Romans or the English. If this assumption be 

 coiTcct it must be a word of comparatively late origin. 



At the same time, the similarity between Cymry and 

 Cimbri — the name of those dread foes of the Romans 

 whom Marius eventually conquered — is so close as to natu 

 rally suggest a common origin for the two name=, if not 

 for the people who bore the names.* The warlike Cimbri 

 have generally been identified with the people who inhabited 

 the Cimbric peninsula, the Chenotusus Cimbrica, now called 

 Jutland. Whether they were connected or not with the 

 Kimmcrioi, who dwelt in the valley of the Danube and in 

 the Tauric Chersonesus or Crimea, is a wider question wnth 

 which we are not at the moment concerned. As to the ethnical 

 relations of the Cimbri, two views have been current, the one 

 regarding them as of Germanic, the other as of Keltic stock. 

 Canon Rawlinson, in summing up the evidence on both sides, 

 believes that the balance of opinion inclines to the Keltic view.= 

 These Cimbri are described, however, as having been tall, blue- 

 eyed, and yellow- or flaxen-haired men. Can we trace anything 

 like these chra-acters in the Cymry ? 



All the evidence which the ethnologist is able to glean from 

 classical writers with respect to the physical characters and 

 ethnical relations of the ancient inhabitants of this country may 

 be put into a nutshell with room to spare. The e.xceeding 

 meaoreness of our data from this source will be admitted by any 

 one who glances over the passages relating to Britain which are 

 collected Tn the "Monumenta Historica Britannica." As to the 

 people in the south, there is the well-known statement in Cresar, 

 that the maritime parts of Britain, the southern parts which he 

 personally visited, were peopled by those who had crosEcd over 

 from the Belga;, for what purpose we need not inquire. Of the 

 Britons of the interior, whom he never saw, he merely repeats a 

 popular tradition which represented them as aborigines." They 

 may, therefore, have been Keltic tribes, akin to the Celti of 

 Gaul, though there is nothing in Ca:sar's words to support such 

 a view. , • 1 



Tacitus, in writing the life of his father in-law, Agncola, 



' Bulletins de In Socifti Anthropologiiiiie de Paris, 2 sir. t. xii. 1878, 



^'^Ibid. t. ix. 1874, P- 662. 3 Hid., t, xii. p. 48S. 



< Prof. Rhys, however, has pointed out that there is no relation between 

 the names. See "British Barrows," by Canon Greenwell and Prof. 

 RoUeston, 1877, p. 632. „ „ ,■ ,v 



5 "On the Ethnography of the Cimbri By Canon Rawlinson. Joiint 

 Anthrop. Inst., vol. vi. 1877, p. 150. See also Dr. Latham s paper and 

 postscript, '■ On the Evidence of a Connection between the Cimbn and the 

 Chersonesus Cimbrica, published in his "Germ.ania of Tacitus." ... 



<> "Britannia: pais interior ab iis incolitur, quos natos m insula ipsi 

 memoria proditum dicunt : maritima pars ab 11s, qui priedx ac belli inferendl 

 causa ex Belgis tr.-uisierant."— ZJc; Bella Gallico, lib. v. c. 12. 



