VOYAGE UP THE COAST TO PARA 147 
clearly marked, however, by an intervening bed of pebbles. 
At Parabyba do Norte the same sheet of drift, but con- 
taining more and larger pebbles, rests above a decomposed 
sandstone somewhat resembling the decomposed rock of 
Pernambuco. In the undecomposed rock below, Mr. Ag- 
assiz found some fossil shells. In the neighborhood of 
Cape St. Roque we came upon sand-dunes resembling 
those of Cape Cod, and wherever we sailed near enough 
to the shore to see the banks distinctly, as was frequently 
the case, the bed of drift below the shifting superficial 
sands above was distinctly noticeable. The difference in 
r 
color between the white sand and the reddish soil beneath 
made it easy to perceive their relations. At Ceara, where 
we landed, Mr. Agassiz had an opportunity of satisfying 
himself of this by closer examination. At Maranham 
the drift is everywhere conspicuous, and at Para equally 
so. This sheet of drift which he has thus followed from 
Rio de Janeiro to the mouth of the Amazons is every- 
where of the same geological constitution. It is always 
a homogeneous clayey paste of a reddish color, containing 
quartz pebbles ; and, whatever be the character of the rock 
in place, whether granite, sandstone, gneiss, or lime, the 
character of the drift never changes or partakes of that 
of the rocks with which it is in contact. This certainly 
proves that, whatever be its origin, it cannot be referred 
. to the localities where it is now found, but must have 
been brought from a distance. Whoever shall track it 
back to the place where this peculiar red soil with its 
constituent elements forms the primitive rock, will have 
solved the problem. I introduce here a letter written 
\)j Mr. Agassiz, a few days later, to the Emperor, which 
will better give his views on the subject. 
