because they spoke Arabic. Nothing could be more erroneous. The material and mental culture of these tribes and their somatic qualities are widely distinct, and the extent of the Arabic language is infinitely larger than the extent of an Arabic racial element. But peninsular Arabia is the least-known land in the world, and large regions of it are even now absolute "terræ incognitæ," so great caution is necessary in forming conclusions, from the measurements of a few dozens of men, concerning the anthropology of a land more than five times as great as France. My own measurements are confined to 38 Annezeh-Bedouins, whom I met in 1883 in Aleppo; 18 other Bedawy, generally Shammar, camel drivers between Mosul and Alexandretta; 20 Mohammedans "Arabs" living in the town Hamah, the site of the first Hittite inscriptions published; and 15 other Mohammedans from Syrian towns. Two groups, unfortunately very small, consist of 6 priests from Gesyra, whom I met in Aleppo, and 5 men from Hail, in Arabia, whom I was able to measure in Constantinople -in all 102 adult men, 61 of them real Bedawy and 41 settled in towns. The cephalic indices of these "Arabs" ran thus: Remarkably parallel with the cephalic index is the form of the nose in both these groups. The Bedawy as a rule have short and fairly broad, the other "Arabs' have, with few exceptions, high and narrow noses, often of an aquiline form. What we generally call a "Jewish type" is found very seldom among real Bedawy and very often among the "Arabs" in the towns, but it would be difficult to reduce this statement to a statistical form, as the conception of "Jewishness" is too uncertain and precarious. We shall later on try to understand the historical connection between these two types, the Bedawy and the other "Arabs." For the moment, we must restrict ourselves to having shown the marked difference that separates them. T. TURKS It is customary in most European languages to call the Mohammedan subjects of the Padishah "Turks." But the word should never be used in this sense without inverted commas; it is more than ambiguous and easily leads to serious misunderstandings. A Turkoman tribe, the Othmanli, commenced from 1289 to conquer a great part of what is now the Ottoman Empire. A good many of the former inhabitants were then forced to speak Turkish and to turn Mohammedans. It is easy to understand that the descendants of the conquerors and of the conquered renegades intermarried freely, and, as the number of the conquering troops was naturally very much smaller than that of the original population, the great bulk of the 10 or 15, or perhaps more, millions of so-called "Turks" has now the physical qualities, not of the conquering Othmanli, but of the old pre-Othmanic inhabitants. So the anthropology of Turkey is, like that of Hungary, a typical example showing how language, religion, nationality, and race are quite distinct conceptions, and it is interesting to see how they are again and again confounded by the general public and by the press. In my paper on the Tahtadji I gave the indices of 187 "Turks" (Turkish-speaking Mohammedans) from Lycia, and was able to show that in the mountain villages, and in some swampy marshes not easy of access, people were generally short headed, and in the towns and on the coast long headed. Since then I have measured 569 more "Turks" from southern Asia Minor and northern Syria, so that I can now publish the cephalic indices of 756 adult men; they run from 69 to 96; if we count the indices 77 to 81 as mesaticephalic, 172 of these 756 men would be dolichocephalic, 151 mesaticephalic, and 433 brachycephalic, with a very pronounced maximum of 77 and 83 men respectively at indices 85 and 86. These numbers speak for themselves, but it is perhaps useful to study first the corresponding figures for the two large remaining groups, the Greeks and the Armenians, and then to compare the results. U. GREEKS What has been said of the "Turks" is valid too in absolutely the same way for the "Greeks" of Anatolia and Syria. Some of them are certainly the direct descendants of old Ionians, Dorians, or Æolians, but the greater part are descended from other groups which spoke Greek and had accepted the orthodox religions. I must here pass over the interesting problem of the Dorian and Ionian wandering and must restrict myself to some measurements taken on a series of 179 adult men calling themselves Greek and belonging to the orthodox church. I published this series in 1890, in my paper on the Tahtadji, and reprint here a graphic table (fig. 1) showing the frequency of the cephalic indices. It is very striking to see how the curve shows a maximum of 22 men with an index of 75, and a second maximum of 18 men with an index of 88. Seventy-nine out of the 179 men are dolicho-, 84 are brachy-, and only 16 are mesaticephalic. If we reckon the arithmetic mean for the whole series, we get an average index of about 80, closely conforming to Weisbach's 95 skulls of Asiatic and European Greeks with an average index of 81.2, and with the series of Klon Stephanos, who found 80.8 for the Greeks in Europe and 80.7 for the Asiatic Greeks. It is easily understood how dangerous and mystifying such an average index may be, if the material is composed of individuals from at least two different groups, as it manifestly is. I am in possession of 93 skulls from a modern Greek cemetery in Adalia; they show about the same distribution of indices. Long before the rediscovery of Mendel and his laws I tried to study the heredity of the cephalic index in the Greek families of Adalia. Here, in the old capital of Pamphylia, there is a large Greek colony, and as I had by good chance been able to give medical help to some of the influential members, I was permitted to measure parents, children and other relations in 67 families. The results were striking. I published a short abstract of them in 1889, in the Reisen in Lykien, and in 1890 in my paper on the Tahtadji. There was a family A; the father had an index of 87, the mother of 73; of the two sons, the elder had an index of 70, the younger 87. In another family, B, the brother of the dead father had an index of 70, the mother 86, a son 82, a daughter 75. In a third family, C, both parents were brachycephalic, with indices of 85 and 86. Of their five children, only the youngest daughter was short headed, with an index of 86, and four elder brothers had long heads with 72, 73, 75, and 73, respectively; 74 was the index of a brother of the mother. If I now study these 67 families in the light of Mendelian researches, it seems as if neither brachy- nor dolichocephaly were dominant or recessive; they seem to be transmitted now with equal frequency, and this has probably been the case for more than 2,000 years. At least, that is the age of the Greek colony of Adalia and for 60 or 70 generations short and long headed "Greeks" have been freely intermarrying. The result was, in many cases, not a mixture, as if we would mix red and white wine, but it was often a manifest reversion to the original types. I called this process "Entmischung,'" but one might perhaps just as well say "Spaltung'" or "reversion" or "restitution.'' जेने Fig. 1. Frequency of cephalic indices in a series of 179 adult male Greeks. 69 70 1 2 3 4 5 6 7 8 9 80 1 2 3 4 5 6 7 8 9 90 1 2 3 4 20 5 In this way good old types, once fixed by long inbreeding, do not necessarily get lost by intermarriage, but often return with astonishing energy. The short heads of the Asiatic "Greeks" certainly correspond to the short heads of the "Turks' and of all the Moslem Sectaries described |