358 Quaternary Deposits, Etc. — Mills. 
A section of the loose materials on the left bank of the Rio 
de Peixe near the Hamlet of San Antonio do Rio de Peixe 
shows from 30 to 40 feet of loess consisting of red clayey ma- 
terial with grains of quartz scattered through it, resting on 
about 10 feet of white and yellow sand, and this on a thin, ir- 
regular sheet of rounded quartz pebbles in which gold and 
diamonds occur. This cascalho is not continuous, but occu- 
pies furrows in the rock, and is no where more than a few 
inches thick. The loess passes by gradation into the underly- 
ing sands. The occurrence of quartz grains in the loess is 
unusual. 
About two miles up-stream from this section, artificial ex- 
cavations afford exposures higher above the level of the 
stream, showing 20 feet in thickness of loess, and under it on 
the rock, thin, irregular patches of gravel in places. 
I did not observe stems of plants or shells or other fossils 
in the loess of Brazil; but I did not make careful search for 
them. My geological studies there were pursued to determine 
the character and extent and value of the gold deposits of the 
region, and I had not time to pursue investigations that did 
not bear on the work in hand. 
Although the region was a low-lying one when the cas- 
calho was deposited, and was probably such when the loess 
was deposited, there is no evidence that the region has been 
covered by the water of either sea or lake during the Quatern- 
ary or recent time. There are no terraces or other effects of 
erosion by waves or currents of a broad sheet of water ; no 
sand ridges or shingle or pebbly beaches or other deposits such 
as are found on shores of seas and lakes, and which are re- 
peated at greater or less intervals on areas over which such 
sheets of water have advanced and retreated. This together 
with the absence of stratification and sorting other than might 
be produced by the streams now flowing precludes the possi- 
bility of submergence of this region during the time of the 
deposition of the loess. The same arguments are valid against 
the theory of the lacustrine or marine origin of loess else- 
where ; but one other condition prevailing here adds great 
weight to the argument; it is the great depth of the softened 
rock underlying the loess. It seems impossible that this thick 
sheet of soft, easily eroded material could be submerged and 
rise again from a sheet of water without being to a large ex- 
