426 FISHER AND HESS [CHAI'. 17 



poorly known. In general the trenches are 1500 to 2000 fm deeper than the 

 nearby sea floor. On the recent cruise of R.V. Argo, a depth of 4935 fm 

 (corrected) was discovered in the North New Hebrides Trench, a depth con- 

 siderably greater than thus far found in any of the other trenches of this area. 



3. Structure of Trenches 



Tw^o major develoj^ments in this century which gave some insight into 

 tectonic processes operating in the Earth's crust were Vening Meinesz' dis- 

 covery in the late 1920's of very large negative gravity anomalies on the 

 convex side of island arcs and in particular over the trenches, and the develop- 

 ment of explosion seismometry at sea by Ewing and Worzel after the war. 

 The prewar explanations for the large negative gravity anomalies were based 

 on an oceanic crust not unlike that of continents but somewhat thinner. Thus 

 a down-buckle of a sial layer provided the mass deficiency required to explain 

 the gravity anomalies. Once the shallow def)th of the Mohorovicic discontinuity 

 under the oceans was established, modification of the older hypotheses became 

 necessary. As a topic this has not yet been thoroughly reconsidered. 



Laboratory experiments by Kuenen (1936) and by Griggs (1939) lent general 

 support to the concepts then current. Both sets of experiments were predicated 

 on a stronger crustal layer floating on a weaker substratum. Kuenen, starting 

 with the assumption of compression in the crust, produced an elastic type of 

 down-buckle to which he gave the name "tectogene". In Griggs' model a 

 viscous liquid substratum was caused to flow in order to simulate mantle 

 convection currents. Viscous drag on the crust produced a downward bulging 

 of the crustal layer in a plastic manner which resulted in configurations over 

 the downward-flowing limb resembling alpine structure. Though much has 

 changed in the years subsequent to Griggs' experiments, there still seems to be 

 a considerable degree of validity in the picture he painted. The earlier structural 

 concepts are summarized by Vening Meinesz (in Vening Meinesz et ah, 1934, 

 1948), J. H. F. Umbgrove (in Vening Meinesz et al, 1934) and Hess (1938, 1948). 



More recently M. Ewing, Worzel, Shurbet, Heezen, Talwani and others from 

 Lamont (Worzel and Ewing, 1954; Worzel and Shurbet, 1955; Ewing and 

 Heezen, 1955), in discussing the Puerto Rico Trench, have discarded the idea 

 that the trenches are compressional features and attribute them to tension 

 (extension). They suppose that the crust was pulled apart and subcrustal 

 material flowed upward into the fissure, giving them their present forms. On 

 both geophysical and geological grounds this hypothesis has weaknesses. 

 Nevertheless, it is interesting to discuss this point of view because it serves to 

 focus attention on certain critical features of trenches. 



The Puerto Rico Trench curves smoothly to the southeast and south around 

 the Virgin Islands. Southward along its axis the Barbados Ridge arises, so that 

 the trench changes into a rise flanked on either side by shallower troughs. The 

 strongly negative gravity anomalies follow the axis from the trench to the axis 

 of the Barbados Ridge, indicating the continuity of the structure and proving 



