346 
Quaternary Deposits, Etc .— Mills. 
tied sands and clays which have been terraced as they rose 
from the sea. The road from Pelotas continues on these de¬ 
posits from 3 to 4 leagues inland, then leaves the terraces and 
passes into a hilly country; and thence all the way to Bage 
and Lavras the rocks are very generally hidden by a deposit 
which is allied in character and I think in method of forma¬ 
tion to the loess of the Mississippi valley and elsewhere. It is 
of fine grain and of even fineness, and shows either no signs or 
very obscure signs of stratification or sorting by water, except 
near the bottom where it sometimes contains streaks of sand, 
or passes downward into sand by gradation. It is generally of 
yellow or drab color except near the surface where it becomes 
colored with organic matter and passes upward into a dark 
colored soil. It is not readily eroded, that is, readily in com¬ 
parison with other kinds of loose materials, and it makes a 
smooth road, not quickly worn into ruts by wagon wheels. 
The loess is not exclusively derived from rocks in the imme¬ 
diate neighborhood in which it occurs, for it preserves iden¬ 
tity or great similarity of character over considerable areas 
within which the rocks vary widely. 
Underlying the loess where both occur, hut also uncovered 
in places along streams and elsewhere is the “ cascalho,” 
which is the gold-bearing loose material of the region. It 
rests directly upon the surface of the rock. It consists of 
gravel and sand or fragments of quartz or other resisting ma¬ 
terial, and is found in two conditions; in one it lies along 
the beds and banks of streams, and has been moved forward a 
greater or less distance by the streams; in the other it rests at 
the outcrop of the quartz vein or other deposit from which it 
is derived. This latter is the debris of the vein or stratum of 
hard, resisting material which, having been left unsupported 
by the wearing away of the more easily decomposed imbed¬ 
ding rock, has become reduced to fragments and fallen along 
where it had existed in place. It is covered for the most part 
by loess, but judging from exposures at streams and elsewhere, 
it occurs in patches over the surface of the rocks generally 
wherever there are veins or strata or other masses of material 
harder than the rocks in which they are imbedded. 
The so-called Lagoa da Nacao is simply a deepened and 
widened portion of the Camaquam river, about 3,280 feet 
long, with an average width of about 84 feet, occupying a por- 
