Status of the Theory of Isostasy. 305 



The strength of the crust is more than enough to sup- 

 port this load, and departures of this amount from per- 

 fect isostasy may be expected over many other parts of 

 the globe. 



Part II is on " Regional distribution of isostatic com- 

 pensation, ' ' the opening portion being an explanation of 

 the geodetic work with the evidence based on the sum of 

 the least squares of the residuals of several solutions, 

 showing that the hypothesis of isostasy brings about a 

 notable agreement between observation and computation 

 of latitude, longitude, and the intensity of gravity. Hay- 

 ford, however, favored the interpretation of local and 

 nearly perfect isostasy. In view of the geologic evidence 

 of regional and imperfect isostasy the data were reex- 

 amined. Two chief arguments were developed. 



First, the maps of the deflection residuals and gravity 

 anomalies show well-defined groupings within which the 

 signs of the outstanding quantities are alike. These 

 groups are unrelated to the vicinity of sea or mountains, 

 and largely, but not wholly, unrelated to geologic prov- 

 inces. They indicate regional areas departing to a nota- 

 ble degree from the form of isostasy given by the 

 hypothesis of uniform distribution of compensation. 

 The diameters of the areas of notable departure run up 

 to about 300 kilometers, 200 miles. 



The second test of local or regional isostasy consisted 

 in taking pairs of adjacent stations, one of which is 

 below the mean level, the other of the pair above. The 

 algebraic difference of the anomalies should be small for 

 the true hypothesis and large for the false. This test 

 eliminates many of the uncertainties which enter into 

 the calculation of the absolute value of gravity for an 

 individual station, and was applied by the writer to the 

 limited amount of data available. Later he learned that 

 G-. R. Putnam, who made important contributions to 

 isostasy in 1894 and 1895, had applied this same test in 

 December, 1912, somewhat more than a year earlier, to 

 Hayford's material, including a pair of stations in Hawaii 

 and another in Switzerland. 14 The number of pairs of 

 suitable stations is limited, but they strongly suggest the 

 existence of regional compensation up to radial distances 

 of at least 167 kilometers. 



14 G. R. Putnam, Condition of the earth's crust, Science, new sei\, 36, 

 869-871, 1912. 



