Lastly there is an idea prevalent in some quarters that at one time there was in Europe a great Celtic empire. Some writers speak of this as though it had been a Gaelic empire. I have been unable as yet to trace this superstition to its source. I suspect that the chapters on Brennius in Geoffrey of Monmouth’s History of the Britons[512] are the real foundation for this strange belief, though naturally no-one to-day would base a serious hypothesis upon so shifty a foundation. M. d’Arbois de Jubainville[513] seems to rely mainly on a passage from Livy[514], in which the writer states that Bellovesus and Sigovesus, nephews of Ambigatus, king of the Bituriges, were sent simultaneously on two expeditions. Livy says nothing of an empire, and the movements which he dates at 600 B.C. seem to have occurred 300 years later. Déchelette[515] had dealt with this absurd notion according to its deserts.
The empire of Ambigatus, if such a thing existed, must have been a Kimric not a Gaelic power. But empires, if we are to understand the word in the sense in which it is ordinarily used, need settled conditions, such as did not prevail in north or north-west Europe until the arrival in the latter region of the pax Romana. It is conceivably possible that among the Kimri the tribal chiefs paid some form of loose allegiance to a super-chief, just as the Dorians, and to some extent the Hellenic world, recognised, very occasionally the hegemony of Sparta; but the evidence which we possess from classical sources does not even imply the existence of any such over-lordship among the Celts. In any case such vague hegemony could only have existed among the Kimric tribes, who for a thousand years harried the people of Celtic lands and the Celtic cradle, Gaelic lords or non-Wiro subjects alike. Before their arrival the Gaelic chiefs ruled only in the mountain zone, and the establishment of an empire in a mountainous country, draining into four rivers and four seas, would have been more impossible than in the open steppe.
APPENDIX I
CHRONOLOGY
BEFORE the days of written history positive chronology is to some extent a matter of speculation, and until the beginning of this century it was little more than guesswork. But the discoveries of Cnossos provided synchronisms between the archæological remains of Egypt and Europe, and since then rival systems have arisen, all of which approximate more or less nearly to the truth. The palæolithic age, however, still remained in the region of guesswork, and wild and very discrepant attempts have been made to estimate its length. It is still the fashion for some writers to use inflated dates and to count years in hundreds of thousands, but the trend of the evidence produced of late is to encourage moderation, and it seems to me possible that the men responsible for the Fox Hall flints, if indeed they are of human workmanship, may not have been separated from their discoverer by a period of time exceeding 150,000 years.
When matters are so problematical, cautious writers are prone to be content with a comparative chronology, or to speak in terms of millennia. This method has advantages, for such writers run little risk of having to confess that they have made miscalculations. On the other hand, the use of actual dates leads to clear thinking, and to gaining a vivid impression of the story, and since we have now good grounds for estimating such dates, (and I shall not be ashamed to own up if later discoveries prove my estimates to be incorrect), I have adopted positive dates throughout, indicating where special uncertainty exists and the direction in which modification may be expected.
While the early palæolithic age is still a hazy past, and the middle palæolithic is not in much better case, the later palæolithic or reindeer age can now be shown to be relatively modern, while the hiatus between that period and the neolithic age has disappeared. Thanks to the work of Baron de Geer[516] we have some foundation for a chronology of this period, and the results of this work have long been made known to English readers by Professor Sollas.[517] There seems to be little doubt but that the pause in the retreat of the Scandinavian ice by Lake Ragunda, which de Geer has dated at 5000 B.C., may be equated, as has been shown by Brooks,[518] with the Daun stadium of Penck.[519] The Fenno-Scandian moraines, on the other hand, can only be equated with the Bühl advance which took place towards the close of Magdalenian times, and this gives us a date of 7000 to 7500 B.C. for Magdalenian. The Goti-glacial moraines seem to indicate the second Würm maximum, and Sollas’ estimate for the interval seems eminently reasonable and has been adopted here; the first maximum of the Würm seems represented by the Dani-glacial line.
The later dates depend, by a series of synchronisms, on those ascertained from the Egyptian monuments, and it is unfortunate that on this point authorities differ. The difference between the various schools of thought has been well and fairly summarised by Dr. Hall;[520] the two great protagonists are Professor Flinders Petrie[521] and Dr. Edouard Meyer,[522] whose system has been adopted with slight modifications by Professor Breasted.[523] For this reason there are alternative systems in vogue for the period preceding 1580 B.C.
Since so many great authorities, well acquainted with the facts and well able to interpret them, differ as to the result, one, who is not an Egyptologist, can decide between them only by testing the application of both systems in his own field of study. Having applied this test to both schemes, I have no hesitation in accepting the latter or shorter chronology, for by the former I find that the earlier periods would be more prolonged than the evolution of the culture warrants. I have therefore, throughout this work used dates based on those given for Egypt by Professor Breasted. This, of course, does not apply to Mesopotamian dates.