FRESH-WATER FISHES OF WESTERN CUBA. 
215 
meridian of Matanzas, and from reports probably beyond Cienfuegos. East of Rincon 
there are, however, frequent river beds, dry during- the time of our visit. This 
main region, belonging to the southern slope, sends a point northward from Rincon to 
the Almendares River in the northern watershed. Aside from the “Ojos de agua” 
along the edge of the cienegas skirting the southern coast there are two notable places 
where undergound rivers find an exit. One at Yento supplies the entire city of 
Iiabana with its water, the other serves to make the region about Guines a garden, 
its waters being used for irrigation. Others in ail probability have a subaqueous 
exit to the south. 
The large spring at Yento, sufficient to supply the city of Habana with its water, 
is the only one which issues on the northern slope, so far as I know. The origin 
of the supply from the Veil to spring has not been traced. It issues but a few feet 
from the Almendares River, and it is very remotely possible, though not at all prob- 
able, that it derives its water from the upper courses of the Almendares. At the 
time of our visit the water of the spring was 1 degree warmer than that of the 
Almendares River. The region north of that river, across the river from Vento, 
being shut out from a possible contributing source, it undoubtedly derives its water 
from the system of underground streams mentioned above. An examination of the 
best available map and the levels of the Western and United Habana Railroads makes 
it seem quite certain that the Vento springs derive their water from the region imme- 
diately south of Vento and north of Rincon and Bejucal — that traversed by the two 
railroads mentioned. This region contains various sinks without surface outlets, as 
well as dry sink-holes, and is the northward-projecting point of the cave region men- 
tioned above. A notable sink-hole in this region is that at Aquada, on the United 
Habana Railroad. This is very broad, shallow, and dry during the dry season, but 
the water rises to stand over 10 feet deep on the railroad track during some of the 
wet seasons. 
THE CAVES VISITED. 
The soil over the region under consideration is thin, the surfaces of the very irreg- 
ularly corroded rocks jutting out in numerous places. This, together with the fact 
that the water of the underground streams is but a few feet underground, gives the 
region an entirely different aspect from the sink-hole and underground-stream region 
in Indiana, Kentucky, etc. It is, in the first place, impossible to enter the under- 
ground streams, and there are no funnels on the surface to indicate the location of 
an underground stream or its tributary. In places the thin limestone roof of an 
underground chamber has given way and enables one to get to the water, which in 
all the numerous places we visited was stationary, not flowing. With one or two 
exceptions the water was covered with a continuous crust of carbonate of lime, due to 
the evaporation and discharge of carbon dioxide from the surface of the perfectly 
quiet lime water. When the water is disturbed flakes of variable size break loose 
and gradually sink to the bottom. All of the so-called caves about Canas, with one 
possible exception, were sink-holes formed by the breaking of the thin roof of a 
larger or smaller underground chamber. In all of the caves where stalagmites and 
stalactites were noticed these extended for 3 feet or more into the water. Inasmuch 
as they could not have formed under water, the latter must have risen since their 
