442 JOSEPH BARRELL 



are not uniform, but in a broad, way are balanced against the relief. 

 This is the proof that a condition prevails of approximate regional 

 isostasy. As the relief of the globe is highly variable, the densities 

 in the lithosphere are therefore within certain limits also highly 

 variable. But on the other hand, the existence of gravity anomalies 

 and deflection residuals indicates that the variations in density are 

 not completely in accord with the demands of the hypothesis which 

 postulates local compensation of the topography uniformly dis- 

 tributed to a uniform depth, nor apparently with any other simple 

 "hypothesis. These quantities measure the differences between 

 the hypothesis and the facts of nature. Let the density variations 

 beyond those required to balance the topography vertically above 

 them be called the outstanding excesses or defects of density and 

 the masses which they represent be called the outstanding masses. 

 It is fundamental to the problems of the strength of the crust, 

 and a system of geologic dynamics in accord with that strength, to 

 determine the depth, form, and weight of these outstanding masses. 

 Do they belong to the centrosphere or to the lithosphere? As 

 Gilbert has noted, if they are to be referred to the centrosphere 

 they do not imply any imperfection of isostasy nor any competence 

 for stress within the crust. Or, if they exist in the lithosphere, the 

 zone of compensation, but are balanced vertically in the same 

 column by other masses of opposite sign, this arrangement will 

 produce local strains within the crust but not tend to flex the crust 

 as a whole. Neither in this case, therefore, would they measure 

 departures from perfect isostasy. As following questions, are the 

 imperfections of isostasy small and local, and the residuals and 

 anomalies the summation of many scattered effects? Or, on the 

 contrary, are there notable regional departures from the conditions 

 of solid flotation which measure a very appreciable rigidity of the 

 crust ? If so, to what extent are these regional outstanding masses 

 related to the mountains, valleys, and deltas in process of evolu- 

 tion under the present cycle of surface activities; producing a 

 progressive unbalancing possibly being slowly restored toward 

 balance by a viscous undertow ? To what extent are the departures 

 from isostasy due to variable composition and density of igneous 

 intrusions dating back to earlier geologic ages, perhaps never in 



