PRESIDENTIAL ADDKESS SECTION A. 43 



square inch to make it begin to yield. After such an experi- 

 ment the granite was found quite as rigid as ever. 



The rocks experimented on, however, lose seriously in rigid- 

 ity when subjected to a high temperature, though they still show 

 great gain in strength with lateral support. 



The evidence of earthquake waves, of earth-tides, and of 

 other phenomena, while firmly establishing the rigidity oi the 

 earth to quick-acting forces, must not keep us frotn giving due 

 weight to the countervailing evidence that distinctly indicates the 

 tendency of a layer lying under the earth's crust to adjust itself, 

 probably with extreme slowness, to the enormous stresses placed 

 on it by the constant wearing away of land surfaces and the de- 

 position of the sediment on the oceanic borders. This combina- 

 tion of rigidity and plasticity is not a singular one. The slow 

 creep of a rigid glacier is a case in point, and probably gives a 

 precise analogy ; or we might cite the case of a copper wire 

 slowly yielding to a heavy load, and yet capable of transmitting 

 longitudinal vibrations. 



Permit me, in closing, to refer shortly to the arguments 

 which practically force us to admit the presence of some such 

 plastic-rigid layer at no great depth, its upper surface lying pro- 

 bably at a depth of between 20 and 100 miles. 



In describing the method of making a geodetic survey, I 

 called attention to the fact that everywhere the plumb-line shows 

 slight deviations, called residual deflections, from perpendicular- 

 ity to the general spheroidal surface. These deviations were first 

 noted by Archdeacon Pratt in calculating the results of the 

 survey of the North of India, and were attributed by him to the 

 attraction oi the Himalayas. On calculating, however, the deflec- 

 tion which the Himalayas undoubtedly produced, he was sur- 

 prised to find that it was much greater than that which was 

 actually observed, and hence that the crust under the Himalayas 

 must be so far deficient in density as to greatly diminish the de- 

 flection produced by the great mountain range. The phenomenon 

 reminded Pratt and Airy, the then Astronomer-Royal, of what 

 would happen if an iceberg were floating near a suspended plumb- 

 bob. Such an iceberg would cause very little deflection of the 

 plumb-line, as its total mass down to the bottom is the same as 

 the mass of the water it displaces. Airy pointed out how strongly 

 Pratt's result bore evidence to the whole crust being, as it were, 

 buoyed up on a liquid substratum, the characteristic of such a 

 substratum being that at the same level it will push up equally 

 and be able to support equal total vertical loads, but would yield 

 if the loads were unequal. 



Within the last ten years Hayford and Bowie, of the United 

 States Coast and Geodetic Sur\'ey, have made this theory the 

 subject of an enormous research. They had exact computations 

 made for every one of 383 geodetic stations in the United States 

 of the deflections which should be produced by all the land eleva- 

 tions and all the oceanic depressions within a distance of 4.126 

 km., or 2,600 miles from each station. They foimd that the mean 



