THE STRENGTH OF THE EARTH'S CRUST 307 



areas of excess of gravity is +0.034. With the exception of the 

 Seattle stations with an anomaly of —0.093, none reach a value 

 of 0.060. It is thus seen that the average notable maximum is 

 not far from twice the mean anomaly. Even by using a uniform 

 divisor of 0.0024 or 0.0030 to convert anomalies, regional depar- 

 tures of load amounting to 1,300 or 1,500 ft. over areas of several 

 square degrees are found to be not uncommon. Over smaller 

 areas the loads rise to about three times the mean, and at Seattle 

 to five times the mean. These figures of course do not measure 

 simply the elevations or depressions of uncompensated erosion 

 features. On the contrary, if the hills and valleys be imagined 

 as smoothed out, then the resulting mean surface would be out of 

 isostatic equilibrium in the same direction over distances amount- 

 ing frequently to hundreds of kilometers and attaining maximum 

 departures too low or too high over smaller areas by these figures. 



But an inspection of the contour map of gravity anomalies 

 (Fig. 5) shows that the large anomalies, those of 0.040 dyne or 

 above, are all located by Hayford and Bowie as centers of maximum 

 anomaly, though the nearest adjacent stations average as much 

 as 100 miles distant. Between the widely spaced stations, the 

 anomaly gradients are gentle. But where the stations form a 

 series closer together, as that from the city of Washington to New 

 York City, the gradients are seen to be steeper and more irregular. 

 It is to be presumed that a further multiplication of stations would 

 show increased complexity over the whole country and reveal 

 maxima higher than those now recorded. The value of the mean 

 anomaly without regard to sign should furthermore increase some- 

 what through the discovery of additional areas of maximum 

 value. Areas of regional positive or negative anomaly would 

 persist in something of their present size, but within broad areas 

 of anomaly of one sign should be discovered smaller areas of oppo- 

 site sign which are now unknown. Upon the completion of such 

 a detailed survey the high anomaly of Seattle would not appear 

 so exceptional as it does at present. 



The chart of the residuals of Solution W shows within the 

 larger areas of like deflection of the vertical many large and sharp 



' Illustration No. 3, Hayford, Supplementary Paper, Bowie, Illustration No. 5. 



