38 R. Ruceies Garzs anp Geo. E. Darsy.—Blood Groups and

and their blood groups taken from different areas vary to an extraordinary degree. Kishi
(1926) recognizes three groups of Ainu, respectively in Hokkaido, Chishima (Kurile Islands)
and Sakhalin, the last two being regarded as purest. By contrast, Furuhata and Kishi (1926)
give the total blood grouping of the Japanese (9,337 tests by various authors, see Table 5),
which they substantiate by 775 further tests in Hokuriku. The record of perhaps the most
isolated Ainu, from the Lu Chu Islands, however, suggests (sce Table 5) that they, too, may
have been originally mostly O.

The striking series of nine native tribes in Formosa (Torii, 1910-1912) do not intermarry
and are now known, from a series of papers chiefly of Japanese origin, to differ markedly also
in their blood groups. The northern Taiyal, who tatoo and have totally different costumes
from other Formosan tribes, have essentially the same blood grouping as the Japanese ; while
in the mountains of the interior the Bunan and especially some of the Tso (Niitaka), bemg more
isolated, have a higher percentage of O, and their uncrossed ancestors may have all been O.
The features of these two tribes retain marked resemblances to those of the Indians. The
Paiwan at the south end of the island, where they show Chinese influence, are high in B and
low in A, as might be expected. The more isolated Tsarisen adjoining them in the interior
are equally high in B and still lower in A, but remarkably high in O, which we regard as the
primitive condition, while the Ami on the east coast are higher in A than B.

It thus appears that in these coastal islands the blood group distributions are consistent
with each other and with the view that this fringe of Asiastic islands contains remnants of a
group of related tribes which remained O after both A and B had arisen and spread through
the continent, and from which the ancestors of the Amerinds were drawn before they were
contaminated with A and (later) B by infiltration from the mainland. ‘T'o throw further ight
on this question, the blood groups of various other peoples, such as the Aleuts, Chukchi,
Ostiaks, Samoyeds, Tibetans and Pagi, should be determined.

Another recent development may have a bearmg on this matter. Thomsen (1930) has
shown that there is another blood group, which is now called Ay. There are therefore four allelo-
morphs, O, Ay, A, and B. The A, receptor differs from A, in giving a weaker reaction. A, is
found to be dominant to O but recessive to A,. There are no intergrades between A, and A,,
but each is a definite entity. It follows that these four allelomorphs which are independently
inherited, will give ten genotypes and six different phenotypes, viz., O, Ay, Ao, B, A,B, ASB:
Later work has shown (Kemp, 1933) that A, and A, can always be distinguished and that
they occur (in European populations)-with a frequency of about 80 per cent. A, and 20 per
cent. Ay. The recently discovered M and N agglutinogens in the blood (Landsteiner and
Levine, 1928) are present in other races, but the percentage of N is found to be lower in Amerinds
than in Europeans. As additional blood reaction tests are discovered it will no doubt be
possible to determine further the relationships between various races on this basis, but the
technique required will become increasingly exacting. Nothing is yet known regarding the
relative frequency of A, and A, in the Australian aborigines and other primitive peoples. The
determination of this may show some general difference between the A of say Australians,
Bushmen and Huropeans.