238 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1916. 



The remark regarding Oldham's results applies also here, namely 

 that it is questionable whether the observations at distances greater 

 than 100° or 120° are sufficiently accurate to justify such definite 

 conclusions. Gutenberg had the advantage, however, of more accu- 

 rate observations than Oldham, and also of measures of amplitudes. 

 There is no a priori reason w^hy the earth might not be made up of 

 a number of shells, but there should be satisfactory evidence for any 

 proposed system; and it must be shown to satisfy the astronomic 

 requirements; or, at least, not to contradict them. Gutenberg's 

 system does not correspond with Wiechert's system of 1897. In the 

 latter a marked change in physical properties occurs at a depth of 

 1,500 km. ; in the former, at a depth of 2,900 km. ; and in crossing 

 into the core, the ratio of the elasticity to the density, according to 

 Gutenberg, rapidly loses six-tenths of its value. This change might 

 be the result of a great increase in density or a great decrease in 

 elasticity; it may be questioned whether the former is compatible 

 with the astronomic requirements, and whether the latter is com- 

 patible with the high rigidity which we know the earth, as a whole, 

 has. So far no answer has been given to these questions. 



In 1879 George and Horace Darwin attempted to determine the 

 rigidity of the earth by measuring the deviation of the vertical under 

 the attraction of the moon. If the earth yielded like a fluid, its 

 surface would always remain at right angles to the vertical, and a 

 pendulum would remain relatively stationary for all positions of the 

 moon ; if the earth were absolutely rigid, the moon's attraction would 

 deflect the pendulum an extremely small amount, but an amount 

 capable of being measured. The Darwins did not obtain definite 

 results because the disturbances of their pendulum were greater than 

 the deflections they attempted to determine. 



A little later von Rebeur-Paschwitz attacked the same problem 

 with better success, using a horizontal pendulum. 



Hecker, in Potsdam, and OrlofF, in Dorpat, have repeated von 

 Rebeur-Paschwitz 's experiment; and both found values for the 

 average rigidity of the earth comparable with that of steel. But, 

 what was most remarkable, and what is still unexplained, the rigidity 

 was apparently greater in an east-west than in a north-south direc- 

 tion. Orloff, experimenting at a gi-eater distance from the ocean, 

 found a smaller difference than Hecker did, and it has been sug- 

 gested that the tides of the ocean are the cause of the difference. 

 The International Seismological Association, at its Manchester 

 meeting in 1911, made plans to repeat the experiments in Paris, in 

 central Canada, in the middle of Southern Africa, and in the middle 

 of Russia ; but no reports have yet come from these stations. 



In the autumn of 1913 Michelson attacked the same problem by 

 a new method, which seems capable of yielding more accurate 



