Status of the Theory of Isostasy. 315 



line, marking the head of navigation in the east, and the 

 valleys of the Pacific coast states. Both large deflections 

 and strong seismicity, however, mean local earth strains, 

 so that such an association is natural. The relationships 

 are worthy of further study. If there are such rela- 

 tionships, they still do not in the least invalidate the 

 evidence bearing on isostasy, as Hobbs's use of the data 

 seems to imply. On the contrary, although the gravita- 

 tive effect of the relief is large in these regions, giving 

 the large topographic deflections, the observations show 

 the actual deflections to be small. The introduction of 

 the hypothesis of isostasy, by offsetting the effect of the 

 relief, brings the calculated deflections into agreement 

 with the observed, and it is therefore just these regions 

 which offer the best proof of the existence of isostasy. 



Hobbs extends his argument (pp. 698-701) to an 

 attempt to invalidate the whole solution of the problem 

 by Hayf ord. His reasoning turns on several points. The 

 most important is that we know nothing concerning the 

 distribution of mass beneath the earth's surface, and that 

 adjacent masses must exert a. preponderant effect. 



To the writer this argument appears to be successfully 

 answered by the results. It is granted that at the begin- 

 ning of the investigation we know nothing of the dis- 

 tribution of mass, since it is the problem to be solved. 

 Introduce a general hypothesis, such as that of isostatic 

 compensation. If, under this hypothesis, the many calcu- 

 lated deflections of the vertical and intensities of gravity 

 agree closely with the measurements, the agreement indi- 

 cates that the hypothesis is a working hypothesis. The 

 hypothesis should then be varied to see how many kinds 

 will survive as working hypotheses and what kinds must 

 meet the fate of the unfit. The working hypotheses, both 

 for deflections and for gravity, are all found to agree in 

 this common feature, — that regional elevations and de- 

 pressions correspond rather closely to regional defects 

 and excesses of density within the outer fiftieth of the 

 earth's radius. The probability that this represents the 

 truth increases with the variety of data tested and with 

 the repeated failure of attempts to establish a successful 

 counter hypothesis. Plausible counter hypotheses have 

 been set up, but like plans for perpetual motion machines, 

 they have not been made to work. Isostasy as a working 

 hypothesis agrees with most of the data very closely, 



Am. Jour. Sci.— Fourth Series, Vol. XL VIII, No. 286.— October, 19 L9. 

 22 



