416 FISHER AND HESS [cHAP. 17 



apparently deepens perceptibly before fading out. Trench-bottom basins 

 commonly extend a few kilometers or tens of kilometers along the axis. Sills 

 sejjarate the basins or ponds from one another, so that commonly the trench 

 axis displays a series of terraces or steps differing in level by several tens or 

 hundreds of meters (Ewing and Heezen, 1955; Fisher, 1961). Modern echo- 

 sounders with high sensitivity, expanded recording scale and precise timing 

 enable one to record accurately reflection-time differences equivalent to depth 

 differences of less than 10 m, even at the greatest depths. Hence, by combining 

 the assumption of sedimentary fill with a plot of multi])lc echoes from soundings 

 taken at few-second intervals, a geologically reasonable and consistent topo- 

 graphic form can be deduced, even when a wide-beam sounder is employed. 

 Isolated, momentary deep returns from an otherwise nearly flat trench liottom 

 are probably spurious ; usually they are due to a variation in electrical input 

 frequency to the recorder drive, or to an especially strong side echo. 



Figs. 1 and 2 are contoured plots of the deepest parts of the Philippine and 

 Marianas trenches, respectively, based on soundings taken in 1959 (R.V. 

 Stranger) employing a depth recorder with precise time base. The Philippine 

 trench plot has been contoured both on first and on strong late arrivals to 

 illustrate the above discussion. Late arrivals suggest that a basin deeper than 

 10,000 m, varying in depth by about 30 m along its mapped axis (reflection times 

 of 12.93-12.97 sec), extends from near 10° 25'N to the bottom of the figure. It 

 is indicated by first arrivals only near 10° lO'N and 10° 23'N. At about 10° 28'N, 

 a sill separates this basin from a similar basin, of slightly shoaler depth, to the 

 north. Reflection times by echo-sounder (vertical lettering) agree with bomb- 

 sounding reflection times (slanted lettering) in the deep parts of the trench. 

 However, generally more echoes can be distinguished on the bomb-records ; 

 greater energy from the explosion results in a detectable echo even from poorly 

 reflecting surfaces like steep trench walls. Positions at the west ends of the 

 Stranger lines were established by visual bearings and radar ranges on islands, 

 and the ship stayed underway throughout the survey. Discrepancies between 

 the trench axial position from this survey and from earlier surveys by Cape 

 Johnson and Galathea are within the limits of navigational errors of the ships. 



Challenger Deep has been investigated recently by Challeyiger (1951), Vitiaz 

 (1957, 1958) and Stranger (1959), and by the bathyscaph Trieste (1960). The 

 contours of Fig. 2 are based on Stranger soundings only. Again discrepancies in 

 location of the greatest depths reported are probably due to uncertainties in 

 fixing the ships' positions. It is likely that all the deep soundings were taken in 

 the same flattish-floored basin. Part of the differences in the Vitiaz and 

 Stranger-Challenger de])ths can be attributed to the velocity correction func- 

 tions used by Del Grosso and Matthews respectively. Also, a variation of 

 recorder-drive input frequency of only 1.7% would account for the range of 

 values reported. Trieste^ depth measurement was made with a pressure gauge ; 

 it agrees well with the sonic sounding. 



It is appropriate here to list the maximum depths reported for 20 of the 

 trenches (Table I). Exploration to date has been sufficiently detailed so that 



