254 ROYAL SOCIETY OF CANADA 
house, and harratta, a wolf, which resemble brithgi, a mongrel dog, 
merch, a woman, awchus, hot, ty, a house, and blaidd, a wolf; but these 
exceptions are few, the bulk of the vocabulary agreeing with those of 
other Dakotan dialects. An instance of purely accidental resemblance 
is that of the Mandan perug, 10, to the Berber forms markoum, mer- 
raoua, 10. All the other Mandan numerals are discordant as compared 
with the Berber, and its perug it shares with the Minetari in peeraga, 
and with the Upsaroka in perakuk. The numerals of the Dakotas in 
general have all their affinities with those of the Ainos of Yesso and 
Saghalien, who seem to have been a timid remnant left behind by the 
main body in its eastern migration. If there be a real connection of the 
two peoples, the modification of the northern Turanian type which the 
Aino presents will account for the apparent traces of European blood in 
the Mandan. The Aino dialects do not furnish the original of warootah, 
a dog, but they give marro for moorse, a wife, yhsehsikka for dsashosh, 
hot, {sise for ote, house, and storkygh for harratta, wolf. There is little 
doubt that the Upsarokas or Absarooke and the Mandans are descen- 
dants of the allied Abhisaras and Nandas, who figure in the ancient 
geography and history of Hindostan; and that they came to America 
from the north-eastern shores of Asia. 
Drake, in his book of the “ North American Indians,” speaks some- 
what contemptuously of the preceding, and other narratives treating of 
Welsh printed books in the possession of Indians, which I have not 
deemed worthy of notice. He says: “There are a great number of 
others who have noticed these Indians; but after an examination of 
them all, I am unable to add much to the above stock of information 
concerning them. Upon the whole, we think it may be pretty safely 
said, that the existence of a race of Welsh about the regions of the 
Missouri does not rest on so good authority as that which has been ad- 
duced to establish the existence of the sea-serpent.” Of course, there 
was no real ethnological science in the days to which these reports be- 
long, nor is there any trustworthy history of the migrations of the more 
obscure Indian tribes. But, the question arises, have the investigators 
of the truth of a wide-spread tradition looked for its confirmation in the 
right quarter or quarters? It is useless to search for descendants of a 
Cymric people among Tuscaroran Iroquois, Algonquin Shawnoes, Musk- 
hogean Choctaws, or Mandan Dakotas; but there are many other families 
of tribes than these whose languages have been more or less investigated. 
Prominent among the Pueblos Indians of New Mexico and Arizona 
are the Moquis and the Zunis. Concerning the former, Dr. Latham, in 
his “Varieties of Man,” cites Mr. Gregg and Lieutenant Emory, two 
