Nov. 21, 1884.] 



♦ KNOWLEDGE ♦ 



■ill 



MAGAZINE OF SCIENCE 



l^iL&INiJWORDED-EXACTDrDESCi^J 



LONDON: FRIDAY, NOV. 21, 1884. 



OONTKNTS OF No. 160. 



FAGB 



Statisti«.-3 of Barataria. II. By 



Grant Allen 415 



Chats about Geometrical Measare- 



mpnt. (Ulna.) By E. A. Proctor il6 

 The Chemistry of Cookery. XLVII. 



By W. Mattieu Williams U" 



The Exploaivenesa of Coal Dust 419 



The Earlh'9 Shape and Motion. By 



Kichard A. Proctor -420 



The Entomology of a Fond. (lUut.) 



By E. A. Butler 421 



Automatic Ventilation, 6o.called, 



(lUus.) 433 



PASS 



First star Lessons. {With 2l>jp^.) 



By Richard A. Proctor 424 



Chapters on Modem Domestic 



Economy. III. (iliiin.) 426 



Reviews : The Antiquity of Man — 



Some Books on our Table 437 



Miscellanea 429 



Correspondence : Some of Your 

 Correspondents — Foreglow — The 

 Weather of 1865 and of 1884, Ac. 430 



The Inventor's Column 432 



Our \\Tiist Colunm 433 



Oar ChesB Coltunn 43-4 



STATISTICS OF BAEATARIA. 



II. 

 By Grant Allen. 



IF we look away from our imaginary instance of Bara- 

 taria, and apply the curious statistical facts we there 

 arrived at to actual cases of mixed population in the real 

 world, we sliall see how very misleading is that kind of 

 ethnology which bases itself entirely on historical or quasi- 

 historical data. Let us take, for example, the case of 

 England. 



People often talk as though the ethnical elements in the 

 population of England must be the same at the present 

 day as they were shortly after the Norman Conquest. It 

 is common to hear thinkers of the purely historical school 

 take it for granted that the proportions which then 

 obtained must still obtain throughout the nation. I believe 

 this to be very far from the real truth. It seems infinitely 

 more probable, in the light of the statistics given by Mr. 

 Galton, that every race is always in a state of perpetual 

 liux ; that large numbers of families are always dying out ; 

 and that other families are always increasing and spreading 

 at an extremely rapid rate. If we could take a glance at 

 the England of the twelfth century, and enumerate all the 

 families it contained, it appears likely that we should ficid 

 ever so many of those families had died out utterly mean- 

 while, whereas the remainder had increased so greatly as 

 to form the bulk of the existing English people in our 

 own day. 



Now, I don't want to press this argument here so as to 

 favour any one particular ethnological theory as to the 

 composition of the modern British nation. I don't believe 

 the time has yet arrived when it would be possible to do 

 this with even the slightest approach to rough accuracy. 

 In order to apply the idea here set forth to concrete 

 ethnology, it would be necessary to make prolonged and 

 systematic research among church registers and other 

 genealogical documents in all parts of the kingdom. But 

 what I want to point out at present is the fact that this 

 particular factor — the relative fertility of special families 

 and special races — is really all-important to the proper 

 study of ethnography. Let us suppose, for example, that 



at any given period the population of Britain consisted 

 one-half of Teutonic Saxons and one-half of indigenous 

 Celts ; then, if the Sa.von women marry on the average at 

 seventeen, and have on the average nine children each, they 

 must have been rapidly increasing ever since ; while if, 

 conversely, the Celtic women marry on ihe average at 

 thirty-two, and have on the average some four children 

 apiece, they must have been steadily decreasing in number 

 ever since. But if, on the contrary (as is more truly the 

 case), the Saxons marry later and the Celts earlier, then 

 the Celts must have been gaining continuously upon the 

 Saxons, and must, in the long run, be gradually supplanting 

 them. 



As a matter of fact, however, in every country where we 

 get varieties of race inhabiting contiguous districts, inter- 

 mixture continuous^ly goes on ; and this intermixture still 

 further increases the difficulty of arriving, by historical 

 research, at any definite result. Still, there is one way in 

 which some approximation may hereafter be obtained, as 

 regards the total of the several proportions, and that is by 

 observation and enumeration of surnames, so far as they 

 can be shown to indicate race. For though a daughter 

 who marries loses her father's surname, and so in mixed 

 marriages merges the marks of her own ancestry in her 

 husband's, yet, since as many mixed marriages are likely in 

 most cases to take place one way as the other (for example, 

 as many Celtic men will marry Teutonic women as Teu- 

 tonic men will marry Celtic women), the women on both 

 sides may be considered to cancel out, and we shall get 

 approximately correct results by reckoning the father's 

 side alone. In short, though a great many Smiths and 

 Browns may be very largely Celtic, yet an equal number 

 of Macphersons, Evanses, and O'Briens are no doubt very 

 largely Teutonic. 



There are, however, a few cases where intermarriage has 

 long taken place on one side only. Take, for example, the 

 Turks in Europe. It is usual to talk of the Turks as 

 Tartars. So far as language and manners are concerned, 

 this may be true enough ; but, ethnologically, it is quite 

 untenable. For years Turks have habitually had in their 

 harems Circassian, Greek, and Slavonic women. Any so- 

 called Turk whose mother was Circassian is only half 

 Tartar ; if his father's mother was also Circassian, then he 

 is only one-fourth Tartar ; if his paternal 'grandfather's 

 mother was also Circassian, then one-eighth ; and so on in 

 a rapidly-vanishing pi oportion of Tartar blood. I say this 

 without political prejudice for or against Turks ; for, 

 so far as I have read, Tartar, Turk, and Circassian are 

 pretty much six of one and half-a-dozen each of the two 

 others. 



It must be remembered, too, that in all times and at all 

 places town population tends relatively to die out rapidly, 

 while rural population tends to roll in upon the towns, and 

 swamp their original ethnical peculiarities. For example, 

 most of the great towns of Britain are situated in what was 

 (during eai-lier ages at least) the most Teutonic^^part of the 

 kingdom. But they are nevertheless saturated through 

 and through with Celtic immigration. Glasgow is full of 

 Highland Scots and Irishmen ; Liverpool and Bristol of 

 Irish and Welsh ; London of all three put together, 

 besides a strong contingent of Cornishmen and half-Celtic 

 Devonians. The rural districts of the Highlands, of Wales, 

 of the west country, of Ireland, are the great breeding 

 places for the modern British races. The people there 

 marry early and rear large families ; part of their surplus 

 population finds an outlet in emigration, and helps to 

 people America, Australia, and the colonies generally ; 

 the remainder rolls back upon the towns, where it inter- 

 marries with the native?, and soon merges in the mass cf 



