THE ISOPIESTIC LEVEL 405 



In this set of computations and also in (2), of the smaller areas and 

 the ocean floors, the sums of the squares of the residuals for the Atlantic 

 ridge depth are much smaller than that for the average depth of the 

 whole Atlantic. This harmonizes with the results obtained on plotting 

 the data (figure 1), so that we may assume that the depth represented by 

 the analyses of the Atlantic island rocks is more approximately that of 

 the ridge on which they stand (about 1,000 fathoms) than that of the 

 Atlantic as a whole. 



The various values for M obtained from the several sets of data are 

 presented in Table VI. 



Table VI. — Isopiestic Depths 



Water-free With water 



. _A_ . A . 



Kilometers Miles Kilometers Miles 



1 56.79 35.29 57.41 35.67 



2 52.64 32.71 52.47 32.60 



3 59.83 37.18 59.25 36.82 



4 55.22 34.31 56.09 34.85 



5 36.20 22.50 33. SI 21.01 



6 96 59.65 



7 60 37.28 



8 40.25 25 



1. Continents and oceans, average Atlantic depth. 



2. Continents and oceans, ridge Atlantic depth. 



3. Small areas and ocean floors, average Atlantic depth. 



4. Small areas and ocean floors, ridge Atlantic depth. 



5. United States area, without ocean floors. 



6. Hayford and Bowie, mountain stations in the United States. 



7. Bowie, stations distributed over the United States. 



8. O. Fisher, estimate by two methods. 



The most striking feature of this table is that all the values are of the 

 same order of magnitude ; and this general agreement may be accepted as 

 evidence that the values approximate to the truth, especially when it is 

 considered that several very distinct methods have been used in arriving 

 at them. The values obtained by the normative method are in very fair 

 agreement, with the exception of that derived from the United States 

 area; the reason for this we shall see presently. The fairly close agree- 

 ment between the four normative values for the isopiestic depth in which 

 the ocean floors are considered and the mountain station value of Hayford 

 and Bowie is a point to which attention may be specially directed. 



As regards the normative values, it would appear that the values for M 

 yielded by the sets of data with large differences in altitude are greater 

 than those yielded by sets in which the differences in altitude are less. 



