Crosby.] 48 [October 2, 
marine agency, but are mountain valleys and gorges, the lower por- 
tions of which have been invaded by the sea; and the bottoms of 
the bays, as already intimated, have in general the forms which this 
origin would require. Not less convincing, as pointed out by Mr. 
Guppy, are the partially filled up valleys of Carenage and Diego- 
martin, the true bottoms of which are undoubtedly below the present 
level of the sea. 
We have no data by which to determine the period when this 
subsidence occurred, except such as is afforded by the degradation 
effected by the sea since reaching its present level; but this is so 
great as to indicate that the existing conditions have been main- 
tained for a long time, perhaps even in a geological sense. The 
shores at most points, and especially where they face the trade wind, 
have the form of bold cliffs forty to four hundred feet high. Owing 
mainly to the moderate southerly dips which the rocks usually 
exhibit, the axis or crest of the range of the Northern Mountains 
lies at most points within from one to two miles of the northern coast; 
and it is interesting to speculate upon the magnificent cliffs which 
this shore will probably present when another mile of land has been 
removed. Not the least noticeable of the coast features of this 
part of Trinidad are the straight, tunnel-like caverns excavated by 
the sea, where it has taken advantage of a softer rock stratum, a 
decomposing mineral vein, or an intersection of well marked joint 
planes. These are well developed about the boca islands, and are 
not uncommonly from one hundred to two hundred feet in length. 
But all these evidences of marine action are confined absolutely to 
the level where the surf beats to-day, and above this line we find 
neither beach, cliff, nor cavern. The bocas are deeper than the 
adjoining seas on either side, being prevented from silting up by the 
powerful tidal currents which set through them. 
It is commonly supposed that the Northern Mountains of Trinidad 
are a direct continuation of the mountains forming the peninsula of 
Paria, and that the bocas are simply depressions in what would be 
otherwise one continuous range. But to my mind nothing is clearer 
than that such is not the case. It needs but.a glance at the map to 
show that the Trinidad mountains do not trend in a direction toward 
the Point of Paria, but to the south of it; while the Parian range 
exhibits an equal divergence toward the north. The two ranges, 
however far extended, would never connect. The boca islands lie 
directly in the course of the Trinidad range, and are plainly a part 
