558 BOWS AND ARROWS IN CENTRAL BRAZIL. 



northern Mato Grosso, and who have shown that the hindrances which 

 the rapids in the streams mentioned opposed to general commerce are 

 not so insurmountable that even great tribal migrations upstream and 

 downstream can not be proved on ethnographic and linguistic grounds. 

 The immense importance which the Mato Grosso possesses for the eth- 

 nology of South America here fully appears, and it follows hence that 

 the knoweldge of the populations of the Mato Grosso must furnish the 

 key for the entire ethnology of South America. 



While the northern border is quite clearly fixed, on the east a limit is 

 less sharply drawn, and the transition to the highlands of Goyaz passes 

 only gradually through separate detached elevations. The sources of 

 the Araguay are to be attributed to the Mato Grosso, while the Tocan- 

 tins belong to the highlands of Goyaz. In the south the Mato Grosso 

 slopes slowly toward the Paraguay basin. Let the boundary be the 

 Serra de Cayapo, which extends from the western edge of the Goyaz 

 plateau in a southwesterly direction to the Paraguay, and on the west 

 side finds its continuation in a range of hills running in a northwesterly 

 direction to the Rio Guapore. The alluvial lowland of the Paraguay is 

 especially not to be reckoned with the Mato Grosso, though in ethno- 

 graphic features it is not easily separated from it. In the southeast, 

 the Mato Grosso is cut off from the Gran Ghaco by the watershed men- 

 tioned, and on the west the Madeira furnishes the natural boundary 

 with its forests, though these begin to appear already east of the 

 Madeira. With exception of the woody river bottoms the Mato Grosso 

 is a pure prairie region, which stretches away between the north flowing 

 river perhaps still farther than the north border of the Mato Grosso. 



It is clear that the Mato Grosso in its central location before men- 

 tioned, endowed with extremely favorable natural conditions, must 

 have played an important role in the history of the South American 

 peoples. Of all the events, however, of which the Mato Grosso was the 

 theater of action nothing more is known. Only from traditional rela- 

 tions of a few tribes or from the narratives of colonists may the latest 

 migrations and invasions be followed. It is therefore not possible from 

 the present condition of knowledge to draw a correct ethnographic 

 picture of the original divisions and dispersions of the populations. 

 We are able from the comparison of materials in museums to gain only 

 a foothold for the knowledge of prior wanderings. The inquiry how far 

 these assumptions can be of use for illuminating the theory of migra- 

 tions of theGes, Tupi, Oarib and Nu-Arawak families by means of com- 

 parative philology, proposed by Yon den Steinen and Ehrenreich, lies 

 outside the borders of this treatise and will be examined later. 



The ethnographic picture of the present Mato Grosso shows, as may 

 be seen from the chart of distribution, a division into four, perhaps 

 three, regions when the arrow or the bow is used as material for com- 

 parison. Each geographic region is characterized by the predominance 

 of a fixed tpye, which is peculiar to one of the above-mentioned ethno- 

 graphic regions. In both classifications the two great areas of east 



