Geography of the Sea. 145 to E. N. E., except at 65 fathoms depth, where there appeared to be an inward flow. On the western side of the passage, near Santo Domingo, the direction of the currents was between S. S. E. and S. W. by W. But few observations could be taken on account of unfavorable weather.
In the Windward Passage, on the western side the currents from the surface down to 130 fathoms set in the directions lying in the S. E. quadrant, and at 200 fathoms the direction changed to W. by S. On the eastern side the surface current varied between E. N. E. and E. S. E., with about ½ knot velocity. Variations in the direction similar in extent characterized also the subsurface currents in the middle and on the eastern side of the passage.
The average of the observations at these three stations gives but a small volume of water passing in either direction.
In the old Bahama Channel, at the station north of Cayo Romano (island off the north coast of Cuba) the currents at and near the surface set south of east; at 65 fathoms, however, the direction varies from about N. W. to E. The deeper current of great volume flowed continually to the north of west with a velocity of over 1½ knots at depths of 130 and 200 fathoms.
Outside the Bahamas, to the north of Great Abaco, a slight current flows about N. W. on the surface and down to 30 fathoms; at 65 fathoms depth the direction changes to a point more westerly, and at 130 fathoms to a point more easterly than the set of the surface current. The maximum in the daily variation at this station occurs about 12h after the moon's transit.
The observations so far as completed by Lieutenant Pillsbury furnish the most valuable data we have at present concerning the Gulf Stream, and it is hoped that further investigation and the analytical treatment of these observations will clearly develop the dynamic laws involved and lead us to a correct theory of current phenomena in general.
Tidal Phenomena.
The causes for many of the inequalities in the tidal elements observed at different places have not yet been satisfactorily explained. The phenomena are dependent on many purely terrestrial conditions. While we are able to ascertain with tolerable accuracy from certain constants, derived from observation, the times and heights of the tides, the problem to compute theoret-
