COASTAL HYDROGRAPHY. 71 



in occasional channels, often only a few feet deep. On the eastern side 

 the gulf is bounded by the present shores of the island. This fiord de- 

 scends farther seaward and joins the Bartlett deep. There are evidences 

 of a similar fiord south of Guantanamo. which is east of Santiago. The 

 channel separating Cuba from the Bahama banks is a valley less than 

 1,800 feet deep in its shallower portions, and consequently no extraordi- 

 nary fiord could be expected. 



Matanzas bay is a land-locked channel three or four miles in length. 

 It is simply the continuation of a partly filled valley, and the bay soon 

 reaches a depth of a thousand feet (1,434 feet at its mouth) before join- 

 ing the outer fiord. Baracoa bay is over 700 feet deep. The valley at 

 Trinidad and many canyons forming outlets for bays are from 100 to 

 200 feet deep. Cabanas and other bays pass into very deep fiords, but 

 often the soundings are only sufficiently numerous to indicate their 

 presence and not their character. 



Many of the bays are from one to ten miles long and form broad basins 

 from which there are very deep though narrow outlets. Thus Havana 

 bay is about a mile and a half wide, with its outlet through a low lime- 

 stone barrier only 900 feet wide and 60 feet deep. Xagua bay (see figure 

 10, page 91) occupies the lower part of a valley 12 miles long and three 

 or four broad. Its depth varies from shoals to 100 feet, but it empties 

 into the sea through a rock-bound canyon 168 feet deep, which in the 

 narrowest portion is only 1,200 feet wide. This bay may be taken as 

 one of the finest types and its origin will be described hereafter. 



Geologic Basement. 



ME TA MORPHIC FORMA TIONS. 



De Castro and Salterain assign the formations of Trinidad mountains 

 to the Paleozoic group, but there appears to be no certainty as to their 

 age. The}^ are composed of a semicrystalline limestone of fine texture 

 and blue color, resembling externally some Ordovician rocks ; so also do 

 the associated black slates. The limestones are somewhat micaceous 

 and are composed of small glassy calcareous particles in a dark matrix, 

 but in their internal structure they are not like our Paleozoic limestones. 

 Some of them become micaceous limestone-schists, so closely resembling 

 mica-schists that there is danger of their being mistaken for them. 



The strata of Trinidad mountains are highly disturbed, being not only 

 uplifted but also bent and overturned, with the beds dipping in variable 

 directions. The modern topographic surface appears independent of the 

 bedding, for the valleys and watercourses which trai^erse the mountain 

 are quite independent of the folds and structure of the strata, but at the 



