554 JOSEPH BARRELL 



thickness, the greatest variations in density take place in the outer 

 shell. This conclusion should be regarded as tentative, however, 

 until confirmed by wider detailed studies and more numerous 

 examples. 



Accepting for the present this tentative conclusion, how does 

 it agree with that previously reached — that isostatic compensation 

 in some regions appears to go notably deeper than 122 km. and that, 

 where deep, the residuals average smaller than for the continent in 

 general ? The answer would appear to be that moderate variations 

 of density are sufficient to account for the isostatic relations of 

 different parts of the continent to each other and that these moder- 

 ate variations may go very deep. 



If the actual distribution of compensation gradually disappears 

 with depth, the hypothesis of uniform compensation complete at a 

 certain depth corresponds to two outstanding masses, one just above 

 the limiting surface, 122 km. in Solution H, the other just below 

 that surface. But these masses would largely balance each other, 

 having opposite signs; so that they would give at the surface of the 

 earth but little evidence of their existence. Imperfections of the 

 hypothesis in regard to the bottom of the zone of compensation 

 would in consequence not readily be detected by methods for 

 determining the depth of outstanding masses. - 



• The isostatic balance of continental crust against oceanic crust 

 is a somewhat different problem from that of the different segments 

 of the continent with respect to each other. Solution H requires 

 a mean difference in specific gravity of about o . i to a depth of 

 122 km. between the crust of the average continental and average 

 oceanic segments. The contrasts in density are therefore pro- 

 nounced and go very deep. Within the continent, on the other 

 hand, the variations in density related to isostatic compensation 

 are comparatively small and this investigation suggests that those 

 variations may be more largely in the higher levels of the crust. 



In conclusion, the depths of the outstanding masses are seen 

 to be related to many problems in crustal statics and dynamics. 

 The depth determines the magnitude of the masses involved and if 

 known will serve as a test of various hypotheses. The excesses 

 and defects of mass departing from that mean which is demanded 



