SECT. 2] TRENCHES 415 



presented arguments and equations based on engineering theory to support 

 his idea that the structures and spacings observed result from an elastic litho- 

 sphere (which overlies a weak magma and is subjected to strong horizontal 

 compression) failing by shear near the continental boundary, with the 

 landward segment overriding and depressing the seaward block to form the 

 trench. Flexure of the lithosphere dies out to either side of the shear plane. He 

 states that on the landward side total stresses on the bottom of the lithosphere 

 are less than those in like position seaward of the trench. Hence, on the land 

 side, highly compressed magmas could be injected into the crust and reach the 

 surface to form the volcanic chain observed. More geophysical measurements of 

 crustal structure, like those recently made in the Guatemala area treated by 

 Gunn, and more data on the physical properties of crust and mantle are needed 

 to test his hypothesis. 



The most spectacular and direct observation of trench-bottom conditions 

 was made in January, 1960, when the bathyscaph Trieste, piloted by Piccard 

 and Walsh, reached the bottom of the Marianas Trench, near Guam, at 35,800 ft, 

 the deepest point in the oceans (Walsh, 1960). They report a soft, gray, sediment 

 bottom, and noted a small shrimp-hke crustacean and one fish at this depth. 



Let us turn to a more detailed discussion of trench topography and structure 

 as revealed by recent bathymetric and geophysical explorations by the authors 

 and others. 



2. Topography of Trenches 



Oceanic trenches are at once among the most appealing and frustrating 

 subjects for bathymetric exploration. To the usual problems of ship positioning, 

 frequency fluctuations and slope returns is added the difficulty of obtaining a 

 recognizable echo from a sea bottom 8 to 10 km beneath the ship. Hence, it is 

 not surprising that nearly every investigation of the very deep trench areas 

 provides data significantly different from that collected by earlier, equally 

 painstaking work in the same areas. 



Methods and special problems of sounding in trenches have been discussed 

 by several authors (Fisher, 1954; Kiilerich, 1959, for example). They pubhshed 

 geometric reconstructions of profiles taken by echo-sounder or bomb-sounding. 

 Such plots demonstrate that rarely does the first echo receive a return from the 

 trench bottom, even when the ship is directly over the trench axis, unless a 

 sounder with an extremely narrow sound beam width is used. Usually weak 

 echoes from one or both trench walls jjrecede the bottom return as the ship 

 moves across the trench. Only by analyzing the entire echo train to resolve the 

 multiple echoes can an equivalent bottom profile be drawn. 



Even the very deep, V-shaped trenches commonly contain a narrow band 

 of nearly fiat bottom, h to 3 km wide, usually attributed to sedimentary fill. 

 This fiat area characteristically gives a relatively strong echo ; on a profile 

 normal to the trench axis its lateral boundaries, the base of the trench walls, 

 can be taken as the points where the strongly reflecting, nearly flat bottom 



