Whether the Cofanes here named are those of the Province of Quitu who murdered the Jesuit missionary, Raphael Ferrer, in 1602, I have not discovered. Perez describes them as still warlike and seclusive, living in the terminal hills of the Cordillera, and avoiding traffic with the tribes of the lower river.[428] An examination of the vocabulary furnished by Saenz inclines me to think that the Churoya may be a mongrel dialect, or at least has borrowed freely from neighboring stocks. I subjoin the principal words from his short vocabulary, with some comparisons: Sun, _mshojaint_. Moon, _juimit_ (_oamito_, Guahiba). Fire, _hijit_ (_chichi_, = sun, Carib). Water, _minta_. Bow, _piranso_ (_paria_, Roucouyenne). Arrow, _funait_. Tobacco, _joo_. Plantain, _parasa_. Dog, _uilg_. Tortoise, _ainjachie_. Wind, _che_. Skin, _begt_. The _Piaroas_ are mentioned by Gilii as a branch of the Salivas, but their language reveals no such connection. They are still found on both banks of the Orinoco above the confluence of the Vichada and near the mouth of the Mataveni. They are savage and superstitious, avoiding contact with the whites; they have had good reason to be extremely distrustful of the advances of their civilized neighbors. They are much given to nocturnal ceremonies, and entertain a great respect for the tapir, who is their reputed ancestor, and also the form which is taken by the souls of the departed.[429] The _Puinavis_ dwell on the Inirida, an affluent of the Guaviare. A tribe, the Guipunabis, is mentioned by Gilii as belonging to the Maipure (Arawak) stock; but it cannot be the same with the one under consideration, the language of which appears to be without affinities. Latham identified them with the _Poignavis_ of the older writers, and on slight linguistic evidence, believed them connected with the Banivas.[430] My own comparisons do not justify this opinion. _8. The Upper Amazonian Basin._ No portion of the linguistic field of South America offers greater confusion than that of the western Amazonian region. The statements are so conflicting, and the tribal changes apparently so rapid, that we are at a loss to bring modern observations into accord with older statements. Thus, I am entirely unable to accept the linguistic classification of Hervas, which certainly was based on the best information of his day. As a matter of comparison I give it. _List of Languages in the Governments of Maynas and the Marañon (Hervas)._ STOCKS. DIALECTS. 1. ANDOA. Araro. Chuudaviño. Gae. Guazago. Murato. Pabo. Pinche. Simigae. Bobonazo.