accurate of the methods now in use. Used in conjunction with the geomagnetic electrokinetograph, it probably provides the best complete picture of the current patterns in the open Gulf. Determination of the flow over the broad, shallow continental shelf remains a difficult problem. Some processes by which the distribution of density is caused to change are evaporation, conduction, and the movement of masses of water by the winds. Since the total transport of water due to the winds in this hemisphere is toward the right of the wind, and since this transport consists of waters in the surface layers which are warm and of low density, the low density waters are piled up at the right of the wind flow, which is in the center of anticyclones, regions of good clear weather. The warm waters are removed from the low pressure storm areas at the left by the wind action. These movements are called the wind-driven currents. Their primary effect is to pile up water of small density in areas of anticyclonic winds and to leave waters of greater density in areas of cyclonic winds. This leads to a secondary effect, namely, the maintenance of a different ocean current related to this distribution of density. Since such currents flow nearly perpendicular to a line connecting the regions having the different water densities, the associated currents form a pattern quite similar to the pattern of the winds. This may readily be recognized on charts showing the distribution of ocean currents with prevailing wirds superimposed. Investigations of ocean currents in the Gulf of Mexico There is probably no part of the oceans of the world of comparable size to the Gulf of Mexico where there is such a wide difference of opinion concerning the specific current regime. This difference is brought out by Sweitzer (1898). He quotes Isaac Vassius who, writing about the year 1663, tells how the currents through the Yucatán Channel "turn obliquely" and pass through the Straits of Florida. The issue of the Encyclopedia Britannica available in 1898 states that "a portion of it (the current-DFL) passes directly to the northeast along the shore of Cuba; but by far the larger part sweeps around the Gulf." Sweitzer himself concludes that, at times "the channel of Yucatán pours its waters into the Gulf so that they spread out in all directions moving on its center," while at other times the currents flow "in a northeasterly direction around the extreme west coast of Cuba." These last results were based upon studies of the distribution of specific gravity of the surface waters, United States Coast and Geodetic Survey, Lindenkohl (1896), and upon modification of currents by the prevailing winds. Sweitzer also reported considerable agitation of the waters covering an area of about 100 square miles occurring off the coast of Texas about 40 miles south and 20 miles east of Aransas Pass which could only be accounted for by the meeting of two opposing currents. Other evidence of converging currents has since been found, and this area has become known as the graveyard of ships. Measurements made in the years 1885 to 1889 by the United States Coast and Geodetic Survey vessel Blake, commanded by Pillsbury (1889), determined the currents in the Straits of Florida. Since the ship was anchored, direct current observations could be compared to computed values, and the comparison provided one of the best examples illustrating the validity of the method for computing relative currents which is now so widely used. Agassiz (1888) published temperature and salinity data collected by the Blake in 1878. These data, together with others collected by the Bache, Bigelow (1917), were used by Wüst to compute the transport of the water through the Florida Straits as 26 million m3/second. Associated with this transport is a water level difference of 19 cm. between the southeastern Gulf and the Atlantic at St. Augustine, Florida, which is discussed by Montgomery (1938). A theory of piling up water in the "Bay of Mexico" was advocated by Benjamin Franklin about 1770. In 1922, the Dana made some observations in the Yucatán Channel and in the Florida Straits, as shown in figure 36. These observations, as well as those of the Mabel Taylor in 1932, were summarized by Parr (1935) who concluded that "evidence thus obtained from the Gulf itself, although directly opposed to some of his premises, nevertheless serves to confirm the theory already advanced by Nielsen on the basis of observations in the Straits alone, that the so-called Gulf Stream only takes the shortest possible path from its entrance through the Yucatán Channel to its exit through the Straits of Florida, without deviating on the way, or diffusing any to itself significant amount into the Gulf of Mexico proper, or receiving any predominant contribution from the Gulf in return." This statement on the one hand and the current pattern shown in figures 34 and 35 on the other hand summarize the present divergence of opinion. The Mabel Taylor cruise was made without unprotected thermometers or other reliable means of determining depth of observations. Parr cautions that particularly in the Yucatán Channel and Florida Straits there is sufficient uncertainty of the depths of the Mabel Taylor operations "to make it seem inadvisable to subject them to a form of analysis and comparison in which depth is an essential consideration." Similarly, the Atlantis cruise of 1934, Parr (1937), lacks subsurface data since the hydrographic cable was lost early in the survey. Thus, the oceanographic data available to Parr were meager. Dietrich (1939) reviewed the currents of the Gulf, and his conclusions, although based upon essentially the same data as used by Parr, show considerably more influence by the Gulf Stream upon the general circulation in the Gulf. He discussed the sill depths showing that the Gulf circulation cannot affect the deep water circulation of the Atlantic below about 800 meters. However, the Florida current, which is shallower than this, has considerable effect. In 1947 the Atlantis conducted a survey of the northwestern Gulf making 27 hydrographic stations (fig. 36) and 473 bathythermograph observations of temperature. These data have been analyzed by Fred B. Phleger (1951), now of the Scripps Institution of Oceanography of the University of California, and have been published by the Geological Society of America. The first cruises of the Alaska, oceanographic research vessel of the Fish and Wildlife Service operating on a survey of the Gulf of Mexico with the cooperation of the Department of Oceanography of Texas Agricultural and Mechanical College and the United States Navy, Office of Naval Research, were completed in October 1951 (fig. 37). These provide the first complete coverage of the Gulf with information needed to compute the deep water currents. The data from these cruises have been distributed and prelimary anal yses indicate that they support the main features of the current pattern shown in figure 34. A brief description of the currents of the Gulf of Mexico is provided in the United States Coast Pilot (1949): Under normal conditions, at all seasons of the year, the great volume of water passing northward through Yucatán Channel into the Gulf of Mexico, spreads out in various directions. Surface flows set westward across Campeche Bank, the Gulf of Campeche, and the Sigsbee Deep; northwestward toward Galveston and Port Arthur; north-northwestward toward the Mississippi Passes; and eastward into the Straits of Florida. A straight line drawn from Buenavista Key, Western Cuba, to the Mississippi Passes forms an approximate boundary between movements having different directions. West of this line the drift is generally northward or westward, while east of it the drift is eastward or southeastward toward the Straits of Florida. There are northward flows along the west side of the Gulf between Tampico and Corpus Christi in the vicinity of the 100-fathom and 1,000-fathom curves, north of the Sigsbee Deep between the 2,000-fathom and the 100fathom curves, and along the west coast of Florida. In general, the surface circulation is the same at all seasons. There is, however, some seasonal change in velocity, the flow being generally stronger in spring and summer than in the autumn and winter. The current near the Florida Keys is variable and uncertain. This description is apparently taken from the Pilot Chart series of the Hydrographic Office (H. O. No. 3500, issued monthly). Another series, H. O. No. 10,690, 1 to 12, Current Charts of the Central American Waters, give resultant direction and velocity for each 1° quadrangle of latitude and longitude. This series has been used by Smith, et al. (1951), to show zones where seasonal convergence or divergence occur. Many of the references cited above contain bibliographies pertinent to the Gulf of Mexico. Also, Geyer (1950) lists many useful works. In summary, the currents of the Gulf of Mexico and their variations are not specifically known. Studies completed in the past indicate some unusual and interesting features and provide. incentive and justification for continued intensive investigation. SEA SURFACE TEMPERATURES A large number of sea surface temperature observations have been collected at shore stations. Some of these data from locations shown in figure |