But there are many reasons, and to me satisfactory ones, for believing that the first Kechuas appeared in South America at the extreme north of the region they later occupied, and that the course of their migration was constantly from north to south. This was also the opinion of the learned Von Tschudi. He traces the early wandering of the Kechua tribes from the vicinity of Quito to the district between the Andes and the upper Marañon, thence in the direction of Huaraz, and so gradually southward, following the inter-Andean plateau, to the northern shore of Lake Titicaca. There they encountered warlike tribes who put a stop to their further progress in that direction until the rise of the Inca dynasty, who pushed their conquests toward the south and west.

The grounds for this opinion are largely linguistic.[282] In his exhaustive analysis of the Kechua language, Von Tschudi found its most archaic forms in the extreme north, in the dialects of Quito and Chinchasuyu. This is also my own impression from the comparison of the northern and southern dialects. For instance, in the Chinchaya (northern), the word for water is yacu, while the southern dialects employ yacu in the sense of “flowing water,” or river, and for water in general adopted the word unu, apparently from the Arawak stock. Now, as Karl von den Steinen argues in a similar instance, we can understand how a river could be called “water,” but not how drinking water could be called “river;” and therefore we must assume that the original sense of yacu was simply “water,” and that the tribes who retained this meaning had the more archaic vocabulary.[283]

Mr. Markham indeed says: “In my opinion there is no sufficient evidence that the people of Quito did speak Quichua previous to the Inca conquest;” and he quotes Cieza de Leon to the effect that at the time of the Spanish conquest they had a tongue of their own.[284] I have, however, shown how untrustworthy Cieza de Leon’s statements are on such subjects; and what is conclusive, there were Kechua-speaking tribes living at the north who never were subjugated by the Incas. Such for instance were the Malabas, whom Stevenson, when visiting that region in 1815, found living in a wild state on San Miguel river, a branch of the Esmeraldas.[285] This is also true, according to the observations of Stübel, of the natives of Tucas de Santiago in the province of Pasto in Ecuador.[286]

This opinion is further supported by a strong consensus of ancient tradition, which, in spite of its vagueness, certainly carries some weight. Many of the southern Kechua tribes referred for their origin to the extreme northwest as known to them, to the ancient city of Lambayeque on the Pacific coast, a locality which, according to Bastian,[287] held a place in their traditions equivalent to that of Culiacan, “the Home of the Ancestors,” in the legendary lore of the Aztecs.

The legends of the ancient Quitus have been preserved in the work of Juan de Velasco, and although they are dismissed with small respect by Markham, I am myself of the opinion that there is both external and internal evidence to justify us in accepting them as at least genuine native productions. They relate that at a remote epoch two Kechua-speaking tribes, the Mantas on the south, and the Caras on the north, occupied the coast from the Gulf of Guayaquil to the Esmeraldas River. The Caras were the elder, and its ancestors had reached that part of the coast in rafts and canoes from some more northern home. For many generations they remained a maritime people, but at length followed up the Esmeraldas and its affluents until they reached the vicinity of Quito, where they developed into a powerful nation under the rule of their scyri, or chiefs. Of these they claimed a dynasty of nineteen previous to the conquest of their territory by the Inca Huayna Capac. They inherited in the male line, and were monogamous to the extent that the issue of only one of their wives could be regarded as legal heirs.[288] They did not bury their dead, as did the southern Kechuas, but placed them on the surface of the soil and constructed a stone mound or tomb, called tola, over the remains, resembling in this the Aymaras.

The extent of the Kechua tongue to the north has not been accurately defined. Under the name Yumbos, or Yumbos de Guerra, the old Relations included various tribes in the Quito region who had not been reduced by the Spanish Conquistadores.[289] A recent traveler, M. André, states that the Yumbos belong to the family of the Quitus, and include the tribes of the Cayapas, Colorados and Mangaches.[290] Of these, the Cayapas and Colorados, as I have shown, belong to the Barbacoa stock, though the term Colorados “painted,” is applied to so many tribes that it is not clear which is meant. The geographer Villavicencio observes that “the Napos, Canelos, Intags, Nanegales and Gualeas, collectively called Yumbos, all speak dialects of the Kechua.” The modern Canelos he describes as a cross between the ancient Yumbos and the Jivaros, to whom they are now neighbors, while the modern Quitos adjoin the Zaparos. Their language, however, he asserts, has retained its purity.[291]

Whether we should include in this stock the Macas, who dwell on the eastern slope of the Andes a few degrees south of the equator, is not clear, as I have found no vocabularies. Velasco refers to them as a part of the Scyra stock, and they are in the Kechua region. Mr. Buckley, who visited them a few years ago, describes them as divided into small tribes, constantly at war with each other. Their weapons are spears and blow-pipes with poisoned arrows. Hunting is their principal business, but they also raise some maize, yucca and tobacco. Polygamy prevails along with the patriarchal system, the son inheriting the property of his father. Some rude pottery is manufactured, and their huts of palm leaves are neatly constructed. Like the Jivaros, they prepare the heads of the dead, and sometimes a man will kill one of his wives if he takes a fancy that her head would look particularly ornamental thus preserved.[292]

The southern limit of the Kechua tongue, before the Spanish conquest, has been variously put by different writers; but I think we can safely adopt Coquimbo, in south latitude 30°, as practically the boundary of the stock. We are informed that in 1593 the priests addressed their congregations in Kechua at this place,[293] and in the same generation the missionary Valdivia names it as the northern limit of the Araucanian.[294] Doubtless, however, it was spoken by outlying colonies as far south as the river Maule, in south latitude 35°, which other writers assign as the limit of the conquests of the Incas.