[2091] Proc. Cambridge Ant. Soc., N. S., iv, 1904, pp. 478-9.
[2092] It has indeed been conjectured, as we have seen (p. 400, n. 3, supra), that the Basques were a distinct race.
[2093] The late Professor F. W. Maitland (Domesday Book and Beyond, p. 222) argues that post-Saxon British survivors could not have been very numerous, as the Celtic language left ‘few traces of itself’; but the same argument might be used to show that when the Romans came to Britain the Celts were few. See F. J. Haverfield, The Romanization of Roman Britain, pp. 9-12.
[2094] An Inaugural Lecture, 1903, pp. 39-40.
[2095] Although Matthew Arnold was almost absolutely ignorant of ethnology, I do not know any book which ethnologists would find more suggestive than his Lectures on Celtic Literature.
[2096] It has been truly said (Journ. Anthr. Soc., 1870, p. xxxvi) that ‘between even the Welshman and the Irishman there is a want of sympathy ... fully equal to that which exists between either ... and the most Teutonic Briton’.
[2097] Mr. Alfred Nutt (Folk-Lore, xv, 1904, p. 234), commenting on a statement in Mr. Nicholson’s Keltic Researches (p. iv) that Buckinghamshire and Hertfordshire are as Celtic (I should say as pre-Celtic) as South Wales and Ulster, asks whether it does not ‘demonstrate the absolute futility of statistics of “relative nigrescence”, or ... size of skulls. The Bucks peasant,’ he continues, ‘may be physiologically akin to the man from Kerry or Glamorganshire; psychically he differs profoundly.’ Yes, but this does not discredit the methods of physical anthropology: it only illustrates what I have said in the text. Between a certain number of individuals in Glamorganshire and a certain number in Buckinghamshire there is, let us assume, physical kinship: if we could isolate those two sets of individuals and compare them, instead of hastily comparing the populations of Glamorganshire and Buckinghamshire as wholes, we might find that the psychical difference was not as profound as Mr. Nutt supposes. Probably it would still be noticeable. But why? Partly because the physical resemblance is combined with a physical difference due to cross-breeding, the degree and nature of which it would be impossible to ascertain; partly because the environment, social, geographical, and climatic, of the peasants of Glamorganshire has for many centuries been very different from that of Buckinghamshire. Let two plum-puddings be made of identical sets of ingredients, but in slightly different quantities, in different kitchens, and by different cooks. The results will be very different. Or suppose that a thousand Spanish immigrants settled in Britain, and intermarried only among themselves. At the end of a century their physical and psychical types would have been modified. Nevertheless, handled with due skill and judgement, statistics of nigrescence and of cranial measurements retain their value.
[2098] See pp. 411-21, supra.
[2099] Geogr., ed. C. Müller and F. Dübner, 1853, p. 948, note to p. 97, line 22.
[2100] See F. Vogel’s ed. of 1888.