ON THE DETEKiMtNATION oK CiHAVJTV AT SEA. 5(Jl 



1. Variation of Gravily ivith Depth. 



In fig. '^ are shown the contours of the ocean floors and the correspond- 

 ing deviations of gravity from the theoretical formula. As the horizontal 

 line represents time, the steepness of the contour is not accurately re- 

 presented. Each circle represents one observation, an open circle indi- 

 cating that the observation is not quite so reliable. The results are to 

 be accepted with caution, for reasons already discussed. Nevertheless, 

 there is a certain consistency about the results which justifies their being 

 brought forward. There is, for example, a well-marked defect of gravity 

 over the Indian Ocean and over its northern extension, the Arabian Sea, 

 and there is a surprising agreement in the contours of the lines of soundings 

 and of gravity, which is particularly noticeable in the part of the voyage 

 from Fremantle to Aden, and would have been more pronounced if the 

 Fremantle observations had been taken as reference points. It is with con- 

 siderable satisfaction that I note a certain measure of agreement between 

 these results and those made by means of the apparatus I have described 

 elsewhere.' The only part of the voyage subjected to a test by this 

 instrument was the approach to Colombo and thence onwards half-way to 

 Bombay. A comparison of the two is shown in fig. 10. The dotted 

 line represents the aneroid results taken from fig. 9, and the black circles 

 the observations made with the ' gravity barometer.' The agreement 

 is not complete, and I have emphasised it by leaving as open circles those 

 which are not in accord. The discrepancy shown by the last three obser- 

 vations may perhaps be accounted for by a break in the thread of mercury 

 which idtimately led to the abandonment of the test {loc. cit.). 



In view of this corroboratory evidence for a fraction of the voyage, I 

 feel justified in venturing upon the following brief discussion of the results 

 (il)tained from the aneroid method, especially as it indicates the type of 

 ])roblem involved in an investigation of this nature. 



Starting from Fremantle the ocean descends to 6000 metres, and 

 gravity falls too ; this defect of gravity is displayed until the island of 

 Ceylon is approached, and indeed continues after the water has got 

 shallow, perhaps due to the influence of the western slope of the moun- 

 tains of which Adam's Peak is the prominent feature. In Colombo 

 Harbour the value is high again, though that port is not much farther 

 from the mountains. On leaving Colombo the depth increases rapidly as 

 the Gulf of Manar is traversed, and gravity falls at the same time in a very 

 remarkable way. The subsequent shoaling as India is approached is 

 accompanied by a rise in gravity, but not quite to the normal value, and 

 there is a persistent defect until Bombay is reached. The suggestion is 

 made that the range of mountains along the Indian coast, the Western 

 Clhiits, is concerned with this defect, which curiously enough reaches its 

 maximum where Mount Hadar 6215 feet, and another 6660 feet, slope 

 down to the coast from summits about 25 miles inland. North of lat. 

 14° 30' the coastal range is less pronounced and tails oft" considerably 

 before reaching Bombay, where gravity regains its normal value (Bombay 

 was taken as a standard station). 



The dip down of the contour into the Arabian Sea coincides with a 

 deficiency in the value of gravity ; the oscillations are probably due to 

 experimental errors, but the mean curve is considerably below the normal 



' ' Apparatus for the Determination of Gravity at Sea.' DuiJield. Boy. Soc. Proc. 

 lyiG o () 



