348 Quaternary Dejyosits, Etc. — Mills. 
one deep furrow I saw it three feet thick. The pebbles are 
partly angular but generally more or less water-worn and 
rounded. The yield of gold of 134 cubic yards taken from the 
cascalho near the outlet was at the rate of 35.3 grains to the 
cubic yard. It was in flattened scales, and evidently had not 
travelled far. Associated with it was black iron-sand, mostly 
titaniferous, but a part of it feebly magnetic. 
A larger part of the surface of the pool was of sand than of 
gravel, and a still larger part was surface of rock. 
There is no loess on the bed of the pool, but it comes down 
to or nearly to the edge of the water on either side, and is 
spread out far and wide over the region generally, thinning 
out and disappearing in places. At one place the excavation 
of the canal near the pool exposed a thickness of 13.6 feet of 
loess. Along the canal the loess in places shows streaks of 
sand near the bottom or passes down by gradation into sand. 
The pool does not occupy the whole of its rocky basin ; but 
the latter extends some distance above the head of the pool, 
and is filled there with sand and gravel brought down, in part 
at least, by the stream. On either side of this portion of the 
basin there are deposits of sand at a higher level than any 
which the water ever reaches now. These sands I did not find 
overlying loess, but they seemed to be continuous with the 
loess, and to have been deposited along the stream at the same 
time that the loess proper was being deposited on either hand 
over the surface of the region generally. 
Evidently the basin was eroded before the loess was deposit- 
ed because sheets of cascalho resting upon the floor of the 
basin pass under the loess ; but above and below the basin the 
Camaquam has considerably lowered its rocky bed by erosion 
since the time of the deposition of the loess, and is still eroding 
it, and this is true of other streams which I saw in Rio Grande 
do Sul. The erosion of so much of the region as I saw, ex- 
cepting in the neighborhood of Cacapava has, however, been 
comparatively slight since the loess was deposited, for the loess, 
although forming but a thin sheet, covers the face of the 
country generally, and the region is one of loess-covered, roll- 
ing or nearly level plains and gentle slopes as a whole with 
small areas of rock exposures. Near Cacapava there seems to 
have been Quaternary or recent local uplifting which has in- 
creased erosion but I had not opportunity to verify this. The 
