Correspondence . 1 9 3 
the water of the upper strata toward the pole, in the latitude of 45°, 
would cause a gradient of about 0.36 of a foot in 100 miles, or about 10 
feet between New York and Brest, in case of a static equilibrium. 
But of course for reasons already given, there would not be really this 
difference — perhaps only about half of it; but this would be sufficient 
to account for the observed differences of sea level between Brest and 
Marseilles, and the gulf of Mexico and New York harbor; the surface 
of the ocean adjacent to the coast of France being about 25 feet above 
mean level, and that adjacent to New York as much below." (While 
I have quoted Science correctly as 25 feet, it is plain that this amount 
should be read as 2.5 feet.) 
This computation it will be seen, is based upon the assumption 
that the gulf stream strikes the coast of Europe with a velocity of 4 
miles in 24 hours. From such data as I have been able to gatherj 
however, it seems certain that the velocity must be at least double the 
above amount, or about 8 miles per day. 
The "current chart" of the north Atlantic, No. 1308 U. S. Hydro- 
graphic Bureau, shows the velocity of the stream ot? the southwest 
coast of Ireland as 8 to 24 miles per day, while in latitude 55°-55° 30' 
W. the velocity of flow is given as 10 to 24 miles per day. 
I therefore conclude that the surface of the ocean adjacent to Eu- 
rope must be at least five feet above, and that adjacent to New York, 
five feet below the mean or normal surface of the geoid. 
Ferrel's position in this matter, as I have previously demonstrated, 
is sustained by the facts of observation, for instance: — 
The gulf of Mexico is 41 inches above the mean sea level at New 
York. (Smithsonian Contrib., 1891, page 196). (Science, Vol. 3. 
page 504.) 
The gulf of Mexico is 3.75 feet higher at the mouth of the Suwanee 
river than is the Atlantic at the mouth of the St. John's river, both of 
these in Florida. (Penny Ency., Vol. 21, page 148, Pearson.) 
The mean sea level of the North sea and the Atlantic at Amsterdam 
and Brest is found to be from 2 to 5 feet above that of the ]Mediter- 
ranean at Marseilles and Trieste, a result, "which is not easy to ex- 
plain on mechanical principles." (Ency. Brit., Vol. 22, page 708.) 
Bourdalone says that the mean level of the Red sea at Suez exceeds 
by 31/4 inches, that of the Mediterranean near Port Said. (Reclus, 
"Ocean," page 2>2>-) 
The Ency. Brit. (Vol. 22, page 708), says these seas are "identical" 
in level, and in another volume it directly contradicts this alleged iden- 
tity of level by saying (Vol. 4, page 792) "We always found a current 
running northward from lake Timseh at the rate of from half a mile to 
a mile an hour." 
This contradiction arises from the fact that the tides in that portion 
of the Mediterranean are known to be but 8 inches, while in the Red 
sea they range from 5 to 6 feet, twice every day; therefore, if the levels 
were identical, the watero of the canal should run towards the south, 
the sea level being then from 2 to 3 feet lower in the Red sea than in 
