201 This assumed sediment velocity value provides a good approximation for the Pacific in the areas of thin sediment cover, but yields erroneously small thicknesses in some types of sediment (e.g. calcareous pelagics) , and in thick or very old sediments (Houtz, 1970) . The time of day and ship's heading appear along the bottom of the profiler sections. The courses shown are courses steered as taken from shipboard logs. These courses generally do not agree precisely with the tabulated

more »

... al data, which are based on the course and speed made good. Each fifth profiler sheet number appears at the top of the pages; the intervening sheets are bracketed by two black dots. Major time-breaks in the profiler records are indicated by slanted lines in the lower time scale. Double horizontal lines appear when the ship's heading changes too frequently to be conveniently annotated. Instrumentation The Navy satellite navigation system (Guier, 1966) was used to obtain frequent and precise fixes. The ship's electromagnetic (E-M) log and gyrocompass were used to interpolate the ship's track between satellite fixes by employing the computer techniques of Talwani (1969). These interpolated ship positions should be generally accurate to better than 0.5 nautical mile, although errors of several nautical miles or more are possible for interpolated positions obtained under extremely unfavorable current conditions. A 12 KHz transducer with an effective half angle of 30° was used with a Times Facsimile recorder for the precision depth measurements. Relative depths can be resolved to about one fathom (1/400 sec of reflection time) in any depth in regions of low to moderate relief. Side echoes are common in areas of high relief and the resolution of small amplitude relief is extremely difficult in such areas. A Graf-Askania seagravimeter (Gss-2) mounted on an Anschutz gyro-stabilized platform was used for all gravity measurements. This system and associated analogue cross-coupling correction devices is described by Talwani (1970) , by Taiwan! et al. (1967), and by Taiwan! and Meijer (this report). The absolute accuracy of the system under normal sea conditions in the open ocean is estimated at about + 3 mgal as discussed by Talwani and Meijer (this report). The relative accuracy of measurements along tracks with constant heading and steady sea-state conditions is somewhat better. A varian proton precession magnetometer was used for all magnetic measurements. The instrument was towed approximately 800 ft. astern of the ship. The accuracy of this type of instrument has been discussed in many publications (e.g. Heirtzler, 1961; Bullard and Mason, 1963) and is generally accepted to be + 10-15 gammas. Data were recorded continuously on an analogue strip chart recorder and also digitally on punched paper tape at a one minute sampling interval. 3 The sound source of the seismic profiling system is a free-firing (410 cm. ) airgun with a repetition rate of about 12 sec. at 140 bars air pressure (Ewing and Zaunere, 1964). The signal is received by a towed hydrophone array, pre-amplifled and fed into a two-channel drum recorder (Ewing and Tirey, 1961). The signal is recorded as variable-density profiles. Data Reduction All bathymetric, gravimetric, magnetic, and navigational data were digitized and reduced with the aid of an IBM 1130 digital computer and on-line Calcomp plotter. The entire data processing procedure including program listings is given in Taiwan ! (1969) and is therefore outlined only briefly. The navigational information (time, latitude, longitude) is first put in digital form and stored in the computer. All time data series (depth, total magnetic field, gravity field) are then reduced in terms of geographic coordinates and distance along the track, by interrogating the digitized navigation, using time as the dependent variable. The precision echo-grams were digitized on a Thomson digitizing table at an irregular sampling interval of 1 to 15 minutes, depending on the topographic relief. The data were then interpolated to regular six minute sampling intervals and the topographic profiles were plotted from the six minute data points. This constitutes about one sounding per nautical mile. More detailed topographic relief cannot be usefully resolved on plots made at the scales presented here. Slope corrections have not been applied and the soundings plotted here have not been corrected for variations of sound velocity in water. The latter corrections are of the order of 100 fm in 3000 fm of water. The bathymetric reduction programs allows the user the option of applying sound velocity corrections according to the Matthews Tables (1939) . The gravity data were introduced into the computer by digitizing analogue records of the sum of gravity beam trace and the calibrated cross-coupling correction. Data points were selected at unequal time intervals frequently enough to describe the analogue trace. The average sampling interval was about 10 minutes. The geographic coordinates, track distances, Eotvos correction, and theoretical gravity (from the International Formula) were then computed, as before, using time as the dependent variable. The gravity system provides an independent check on the inferred east-west velocity of ship (i.e. Eotvos effect). Considering this factor, careful editing of the navigation was done in arriving at the final, internally consistent navigation. The shipboard observations were tied to known land-based gravity base stations at each port. The magnetometer data were introduced into the computer by either digitizing the analogue records or by using the carefully edited digital punched tape-depending on the quality of the respective records. The geographic coordinates and track distances were computed for each sampling point, again using the navigational data and time as the dependent variable. The regional field at each navigational point was determined, using a modified computer program of the Goddard Space Flight Center, which describes the regional field by spherical harmonics of degree and order 11 (Cain et al., 1964). The regional field at each geomagnetic sample point was then obtained by linear interpolation of the field determined at navigation points. Since navigation points were rarely separated by more than a few hours (or a few tens of miles) and since spatial changes of the regional field are small over short distances, this approach introduces no significant error in the determination of the regional field at each individual sample point. No indications have been given regarding magnetically quiet or disturbed days and no Lamont-Doherty Geological Observatory Personnel -Geophysics Party Aboard U.S.N.S. ELTANIN (Gruises 28-32)