430 



FISHER AND HESS 



[chap. 17 



account for the mass distribution, the seismic velocities, and the forces acting 

 to keep trendies far out of isostatic equihbrium. Such a hypothesis is neces- 

 sarily lii<ihly speculative but it is at least better than no reasonable hypothesis 

 — which is the j)resent status with respect to this problem. 



The forces necessary to produce the structure and departure from isostatic 

 equilibrium will be attributed to mantle convection currents. The only other 

 source which has been suggested, thermal contraction of the earth, seems the 

 less likely of the two at this time. The supposed nature of the structure is shown 

 in Fig. !>. It differs from previous proposals in that the mantle and "crust" 



Consolidated 

 sediments ond 

 volcanic debris 



1000 + kg /cm 

 i 



T I 1 1 1 1 I I r 



deserpentinization 

 + 500°C 



Partial fusion 

 mogmo generation 



Fig. 9. Supposed structure, with typical seismic velocities (l'^ in km/sec), for a hj])0- 

 thetical trench-island arc association. 



are moving together as a unit rather than as in Griggs' (1939) experiment where 

 the weaker mantle material was moving at a higher velocity and acting on the 

 crust by viscous drag. It is postulated that the "crust" is serpentinized peridotite 

 (70% serpentinized, Vp = Q.l km/sec, density 2.86 g/cm^) rather than basalt. 

 The crustal layer goes down with the descending limb of the convection cell 

 until it reaches a temperature in the neighborhood of 500°C where a 

 deseq)entinizing reaction takes place, releasing water. Similarly, the original 

 supracrustal volcanics and sediments descend until melting or partial melting 

 occurs. Fluids, magma or water, rise, migrating toward the island arc or 

 concave side of the structure. The reason for migration upwards on the con- 

 cave rather than the convex side of the arc is given in Fig. 10. On this side 

 there will be a tendency for open fractures to form at right angles to the direc- 

 tion of compression (at right angles to the trench axis). Strong deformation 



