284 GROVE KARL GILBERT -DAVIS 



earth, occur at continental borders, because land gravity is there increased by the excess of 

 attraction of the dense suboceanic crust and ocean gravity is there diminished by the defect of 

 attraction of the lighter subcontinental crust. Attention is next directed to the Appalachian 

 belt south of the area of glaciation, where it "has been practically free from loading during two 

 geologic ages, and has been unloaded to the extent of many thousands of feet of rock." A 

 lagging of readjustment by underflow should here result in a defect of gravity; yet "there are 

 large plus anomalies in that belt, and their existence evidently creates a difficulty in interpreting 

 plus anomalies as due to crustal excess of mass"; for "isostatic adjustment may be supposed 

 to have added mass in compensation for the unloading, or for part of it, but may not plausibly 

 be supposed to have overcompensated so as to create an excess of mass." In the third place, 

 the Delta of the Mississippi is instanced as a district of great loading in recent time; "and if 

 isostatic adjustment by outward underflow has not kept pace with the loading there should be 

 in this district a local excess of mass. The fact that the local anomaly is minus instead of plus 

 calls in question the mode of interpretation which infers defects of mass from minus anomalies"; 

 and this must be accepted because here the Helmert effect is in some way masked. 



In review, it is concluded that, although an approximate isostatic equilibrium has been 

 shown to exist and although it is therefore appropriate to search for processes that produce and 

 maintain it, underflow driven by gravity must hold a low rank among these processes, because 

 the gravity anomalies shown on the map are frequently the reverse of those that should occur if 

 underflow held the high rank that has often been given to it. Nevertheless the Laurentian 

 region is noted as having subsided when it was heavily covered with ice in the Glacial period, 

 and as having risen when the ice melted away, as if the disturbance of equilibrium by loading and 

 unloading at the surface had there been rather promptty compensated by outflow and return 

 flow deep beneath the surface ; and this would seem to contradict the conclusion just announced, 

 and to show that gravitative underflow ought to hold the relatively high rank as a means of 

 producing and maintaining crustal equilibrium that has commonly been ascribed to it. But 

 in answer to this one may suggest that the Laurentian region, having been long untroubled by 

 forces of deformation, has attained so delicate a balance that it responded quickly to the 

 accumulation and disappearance of the great ice sheet, which appears indeed to have been thick 

 enough to enforce its demands; while recently deformed regions which, hke the upthrust high- 

 lands of the Rocky Mountains, maintain a considerable overload of rock may be still so con- 

 strained by the upthrusting forces that, much as they would hke to settle down in response to 

 the overload, they are not allowed to do so. However this may be, Gilbert's conclusion was 

 that the gravity anomalies of regions that have been unloaded and loaded at the surface by 

 erosion and deposition are not such as would result from a lag in the underflow by which new 

 matter would be introduced beneath the unloaded region and withdrawn from the loaded region. 

 Some other cause for such anomalies must be found; and Gilbert's belief was that the cause 

 lies in "irregularities in the vertical density gradient," and that this cause "appears both quali- 

 tatively and quantitatively competent." The argument by which this conclusion is reached 

 is an excellent example of Gilbert's method of thinking and deserves close study. 



