[CAMPBELL ] MEXICAN COLONIES TRACED BY LANGUAGE 257 
okpi, the sun, and the Kulanapan and Yuman poh and aawoh, fire, are 
purely Berber, atfuct and aphougo. The Yuman echamay, brother, is 
not modern Cymric, but it is the Berber ygooma; and its jacuel, child, is 
nearer the Berber ayal than the Gaelic gille. The Yuman eye is 
edotche; the Pujunan, il hil, and the Berber dialects represent these by 
thitta and elu. To eat and to drink in Yuman are asao and hasue, in 
Berber, ichchi and iswa. Earth is amut in Yuman, and in Berber 
tamouts, because the cis-Atlantic Celts seem to have lost the Berber and 
Gaelic usage of a merely euphonic initial and final t. Dog is ahut in 
Yuman, and horse is huts; these in Berber are aidi and aoude. The 
Kulanapan kaiyah, the head, dahhats, a girl, dahno, a mountain, dah, 
a woman, can, 2, tsadi, 6, are the Berber agaio, thagshishth, thener, adi, 
seen, and sedis. But the vocabulary must speak for itself. What it 
does say is not in favour of Madoc’s mythical Welsh colony, but of a 
Cymric Berber offshoot allied to yet differentiated from the Peruvian 
stock. 
There is no evidence that the Yumans, Pujunans, and Kulanapans 
were ever in Peru. From their present geographical position, it is more 
natural to regard them as colonists from Mexico. The Coco-Maricopas, 
and this will include the whole Yuman family, are regarded as intruders 
in their present habitat. If they came from the south, that is, from 
Mexico, as a remnant of the Olmecs, whose main body went to Peru, 
the tradition of the Welsh Indians falls to the ground. But the fact 
has been established beyond doubt, that there are Cymric Indians both 
in North and in South America. Tradition says that such were found, 
first in the Carolinas, and afterwards far up the Red river beyond the 
Mississippi, in other words due west of South Carolina, and tending to- 
wards New Mexico and Arizona. They were plainly a people of a reced- 
ing area, driven westward by stronger and more numerous tribes. This 
leads to the belief that part of the original colony from the Canary 
islands landed on the coast of Florida, while the rest went on across the 
gulf to Potonchan in Mexico. The eastern section must have moved 
northward into what became the Carolinas, and, sometime after the 
second half of the seventeenth century, have migrated westward until 
they settled in western Arizona and Southern California. I have looked 
for traces of their speech through vocabularies of various south-eastern 
families, the Caddoan, including Pawnees and Ricarees, the Adaizan, _ 
Uchean, Attacapan, Apelusan, Shetimashan, Timucuan, Catawban, and 
Nachesan, but without satisfactory results. It is, however, a curious 
coincidence that the Adaize numeral 8, which is pacalcon, should be so 
like the Peruvian Aymara and Sapibocono 7, which is pacalco and 
