of sea water in the different parts of the ocean, &c. 175 
to the general fact which I have stated above, and stands in 
need of explanation. 
In order to account for this, it has been argued that the 
Mediterranean sea is not supplied by the rivers which flow 
into it with a quantity of fresh water, sufficient to replace 
that which it loses by evaporation under a burning sun, as- 
sisted by a powerful radiation from the African shores, and 
the parching winds blowing from the adjacent deserts ; so 
that a current from the ocean is required to replenish this 
waste, and prevent the Mediterranean from sinking below its 
level. Accordingly it is observed that a current sets in at all 
times from the ocean into the Mediterranean, which current, 
I am informed, is so strong at Gibraltar, as to carry a ship at 
the rate of two or three miles an hour; and it is felt as far 
eastward as the Cap de Gat, a distance of upwards of one 
hundred and fifty miles ; so that ships going out of the Me- 
diterranean, scarcely ever attempt to beat out against contrary 
winds, and usually keep close either to the African or Euro- 
pean shore, in order to avoid the full force of the stream. 
If this hypothesis, however, of a disproportion between the 
loss of water produced by evaporation, and the .inadequate 
compensation afforded by the ingress of rivers, be founded in 
fact ; and if this deficiency be replenished by the saline waters 
of the ocean, it will be necessary to explain why the waters 
of the Mediterranean do not gradually increase in saltness, 
and indeed how it happens that they are not ultimately con- 
verted into a saturated brine. It has been supposed, in order 
to remove this difficulty, that an under-current of water, salter 
than the ocean, runs out of the Mediterranean at the Straits 
