140 BRAZILIAN ROCK INSCRIPTIONS. 
I sent one of my assistants, Mr. Barnard, to examine a burial 
station on the Island of Marajó, and he brought me a small collec- 
tion.of pottery presenting some interesting features. In this arti- 
cle I shall confine myself to a description of the inscriptions I 
have collected, hoping in another article to describe the pottery 
and other relics. 
The Tocantins inscriptions occur at Alcobaça, a point on the 
left bank of the river, near the first falls, and about one hundred 
miles from the mouth of the river. Here are exposed on the 
banks during the dry season beds of a fine-grained, very hard, 
dark red or brown quartzite, the strata having only a slight dip. 
Joints divide the beds into large blocks which often lie in place, 
but along a part of the shore they are piled up in confusion. Dur- 
ing several months of the year, when the river is high, the locality 
is under water, as is the case with similar incised rocks at Serpa 
on the Amazonas. My guide told me that here were letreiros, oF 
Indian inscriptions, and I was fortunate enough, not only to find 
several, but to be able to bring away with me two small incised 
blocks. The figures are pecked into the rock by means of some 
blunt pointed instrument. They are so rude and irregular, that I 
see no reason why a pointed stone may not have answered the pur- 
pose. The grooves are usually wide and not very deep. Occa- 
sionally the unskilful hand missed its mark and marred the figure. 
ese figures are usually cut on the sides of the blocks of 
rock and show much wear; many are hard to trace, and the 
majority are more or less covered by a shining black film of 
manganese deposited by the water. The surface of one of my 
specimens, Pl. 2, fig. 5, has a metallic lustre, like that of a well 
blackened stove. 
Of these inscriptions, Pl. 2, fig. 1, which is about sixteen inches 
in length and is somewhat badly preserved, appears to represent a 
human figure, but it has a decapitated look. It may perhaps be 
intended to represent some lower animal. The position of the 
arms and legs conforms to the type of ordinary Indian representa- 
tions of the human form, as we shall see further on. 
The other figures are, for the most part, more or less complicated 
spirals. Pl. 2, figures 2, 4, 5,7, and 11. One of these, Pl. 2, fig- 4, 
may represent the human face, the upper diverging lines being the 
eyebrows, the medial descending loop the nose, and the spiral the 
eyes. Equally rude representations of the face occur elsewhere- 
