356 H. H. ROBINSON 



involves a simpler assumption respecting the distribution of stresses 

 within the earth's crust, and for this reason should be considered pref- 

 erable to the hypothesis of differential tilting. 



If earth movements in the Great Lakes region had been great 

 instead of small, and the errors in the water-plane determinations 

 small instead of great, the character of the crustal deformations might 

 well be ascertained; but on the contrary we find the probable error of 

 observation ?o large — it is expressed in feet, rather than tens of feet or 

 in inches — that it leads to the impression that but little refinement of 

 the old idea of differential warping is possible on the basis of the 

 ancient shore-lines alone. 



