Chap. VI. EIVER BEEVES. 221 



palms, which, however, were visible only as isolated 

 groups of columns, tufted at the top, rising here and 

 there amidst the waste of waters. In the afternoon we 

 rounded the westernmost point ; the land, which is not 

 terra firma, but simply a group of large islands forming 

 a portion of the Tocantins delta, was then about three 

 miles distant. 



On the following day (25th) we sailed towards the 

 west, along the upper portion of the Para estuary, which 

 extends seventy miles beyond the mouth of the Tocan- 

 tins. It varies in width from three to five miles, but 

 broadens rapidly near its termination, where it is eight 

 or nine miles wide. The northern shore is formed by 

 the island of Marajo, and is slightly elevated and rocky 

 in some parts. A series of islands conceals the southern 

 shore from view most part of the way. The whole 

 country, mainland and islands is covered with forest. 

 We had a good wind all day, and about 7 'p.m.. entered 

 the narrow river of Breves, which commences abruptly 

 the extensive labyrinth of channels that connect the 

 Para with the Amazons. The sudden termination of 

 the Para at a point where it expands to so great a 

 breadth is remarkable ; the water, however, is very 

 shallow over the greater portion of the expanse. I 

 noticed, both on this and on the three subsequent occa- 

 sions of passing this place in ascending and descending 

 the river, that the flow of the tide from the east along the 

 estuary, as well as up the Breves, was very strong. This 

 seems sufficient to prove that no considerable volume of 

 water passes by this medium from the Amazons to the 

 Para, and that the opinion of those geogi'aphers is an 



