776 C. B. BROWN ON THE ANCIENT 
these cases sections of the deposit would show their origin—the delta- 
beds by their containing brackish-water shells, and the lake-beds by 
their even and regular deposition. None of the beds of the deposit 
now under consideration have been formed thus. 
Where there are no old deposits left on either side of a valley 
we know that the river, by cutting close up to the underlying rocks 
on either hand, has taken away on its return that which it laid down 
previously ; and then we find it bordered only by recent alluvium. 
This will be seen where a river flows swiftly along a narrow moun- 
tain valley, but not where it winds along a sloping plain. 
In some instances, where the river suddenly abandons its old 
course by cutting a new one across a neck of land between two 
curves, it leaves the old channel as a lake to be filled up very slowly 
with fine materials. Thus irregularities in the surface of its older 
deposit were produced, causing it in places to vary very much in 
altitude and thickness. Good examples of these old river-courses are 
to be seen in its present alluvial plain at Monte-Alegré great lake, 
at Jamunda-mouth, Sapukia, and other places. 
On the Purus and Jurua, where there are numbers of these river- 
lakes, there are instances of the rivers having cut through narrow 
necks, leaving large loop-shaped patches of still water. The ex- 
istence of these loops explains a difficulty which it would otherwise 
be hard to account for. I refer to the river passing from one side 
of its valley to the other, and leaving a portion of an older for- 
mation untouched at a higher level than its plain. We have a case 
of this sort in a patch of Tertiary clays at St. Paulo, the river having 
at one time flowed on the south side of it, while now it is on the 
north. ‘To account for this, we must look upon such a tract as the 
centre portion of a loop, which the river partially flooded in the 
seasons of high river. 
The recent alluvium on the Purus river is a grey clay upon grey 
sand; but in places where the river has cut through one of its old 
silted-up curves, left originally as a sort of lake, it is a grey loam, 
containing great quantities of unaltered leaves and tree-stems in 
thick beds. The alluvium of the Solimées is also a grey clay, upon 
which is a brownish loam; and the material being now added to its 
surface in flood-time is a greenish-brown arenaceous loam. Upon 
the old deposit this latter is represented by the red loam so fre- 
quently mentioned in the foregoing sections, which must have been 
thrown down upon the sand and clay beds in still water during 
times of flood. It is very persistent in its occurrence over the 
whole valley, and appears on the surface of the old deposit at various 
levels, from having settled down in some places in shallows, and 
in others in deep water, corresponding to the present alluvial 
lakes. 
The surface of the great valley-deposit has of course been modi- 
fied to some extent by atmospheric agencies, and is much cut up by 
small streams, which have formed gullies and ravines through it. 
Upon the accompanying map (Pl. XX XVIII.) I haye traced out 
the approximate boundaries of the old deposit and the recent allu- 
