SECT. 1] GBAVITY AT SEA 135 



accuracy better than 2 mgal would be useful only in a limited number of cases 

 where exceptionally accurate navigation is possible. The amplitudes of the 

 horizontal- and vertical-wave accelerations experienced on a surface ship usually 

 lie between 10,000 and 100,000 mgal depending on the state of the sea and the 

 size of the ship, but these accelerations decrease with depth below the surface 

 and are typically less than 5000 mgal at 250 ft depth. 



The first attempts to measure gravity at sea were made by O. Hecker, and later 

 by W. G. Duffield, using a barometric apparatus on board a surface ship. 

 Although an apparatus of this type constructed by H. Haalck eventually gave 

 results of useful accuracy in extremely calm water in 1938, this method has 

 never been widely used. 



From the first cruise of the Dutch submarine, K II, in 1923, up to 1955, all 

 reliable measurements of gravity at sea had been made with the Vening 

 Meinesz pendulum apparatus operating in a submarine. This apparatus has 

 been extensively described in the literature (Vening Meinesz, 1929, 1941). It 

 is a thoroughly reliable apparatus, if handled with due care, and the accuracy 

 of a measurement is between 1 and 2 mgal when the wave accelerations are less 

 than about 6000 mgal and when there is little energy present with the same 

 period as the pendulums. These conditions can normally be realized with a 

 submarine able to dive to 250 ft. A total of some 4214 measurements have been 

 taken with this apparatus, of which 2943 were taken by the Lamont Geological 

 Observatory of Columbia University. The apparatus has the disadvantage that 

 the computation of the gravity values from the photographic records is lab- 

 orious, taking about one hour per record. It is also not suited for development 

 into a surface-shi]3 instrument. 



Gravity meters have replaced the pendulum apparatus for measurements on 

 land with the exception of absolute determinations of gravity and the establish- 

 ment of calibration ranges. The first gravity meter used at sea was the vibrating 

 string instrument constructed by R. L. G. Gilbert, which was successfully 

 tested on board a submarine in 1948 (Gilbert, 1949). Unfortunately, it has not 

 been used successfully at sea since these tests, although a somewhat similar 

 meter, developed from the Vibrotron pressure gauge by the Borg-Warner 

 Corporation, was successfully tested on a surface shi]3 in 1959. 



In the meantime L. J. B. LaCoste and Anton Graf had independently 

 develojDed spring-tyxDC gravity meters capable of functioning on submarines. 

 The LaCoste meter was tested by F. N. Sf)iess and G. L. Brown on two cruises 

 in 1955 (Spiess and Brown, 1958). Sixteen previously established pendulum 

 stations were reoccupied on the first cruise and on the second they made 15 

 simultaneous pendulum and gravity-meter measurements. The root-mean- 

 square difference between the pendulum and meter observations was 3.3 mgal 

 in both tests, which indicates that the standard error of a gravity-meter 

 reading is about the same as that of a pendulum observation. The first tests of 

 the Graf meter were made on a surface ship on Lake Starnberg in 1 955 (Graf, 

 1956). Worzel and Graf (see Worzel and Graf, 1957) tested Graf's instrument 

 by comx^arison with simultaneous pendulum observations at 53 sea stations 



