144 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
Downstream the form of the channel is that of a depressed v-shaped 
valley, and this finally merges into the broad, open, but flat-bottomed 
valley, with the high banks far apart. 
The Amazon is also too powerful a stream to be shut up by its own 
silts. 
Evidence of the islands. — The process by which islands are formed 
by depression consists in the lowering of the general land surface until 
the sea passes up some of the valleys and gets behind or landward of 
some of the highest points. The island of Itamaracá, for instance, before 
it became an island, had one stream flowing past its northern end — the 
Tijucapopo (or Catuáma) — and one flowing past its south end — the 
Iguassú — and the canal that now separates it from the mainland was a 
valley through which small streams drained both north and south. 
When the depression of the coast took place the water backed up into 
the lower parts of these streams and into the valley behind it, thus 
entirely cutting it off from the mainland and leaving it an island. 
The island of Itaparica and the other islands in the Bay of Bahia are 
also the higher portions of irregular land surfaces let down by a depres- 
sion till the water of the sea flowed round them. Пара са channel 
is a drowned valley that formerly drained northward into what is now 
the Bahia de Todos os Santos, but which was formerly a valley. The 
same thing is true of the islands of Tinharé and Boipéba on the coast 
south of Bahia. 
If we wish to test the theory of these islands having been made as 
here suggested we have but to imagine the effect of an elevation upon 
them. The deepest water recorded in the channel west of Itamaracá is 
less than three fathoms. If the land were elevated more than three 
fathoms evidently this water would run out and we should have a small 
stream flowing where tide water now enters. Many islands on a coast 
are usually regarded as evidence of a late depression of that coast. It 
might be supposed, then, that inasmuch as the Brazilian coast has re- 
markably few islands, this absence of them might be taken as evidence 
against a late depression. The islands that do exist, however, bear out 
the theory of a depression, while the absence of a large number of small 
outstanding islands bears out the idea that the depression took place 
long enough ago to allow the sea to obliterate them, and in some in- 
stances to throw their remains on the shores of the mainland. These 
same silts also helped join to the mainland some of the in-shore islands, 
Off-ahore clays. — The borings made by Sir John Hawkshaw on and be- 
hind the reef at Pernambuco in 1874 penetrated to a maximum depth of 
