RIGIDITY OF THE EARTH 



103 



It was found that there was a gradual downward trend in the 

 E.-W. curve during August, but Httle or none in the October- 

 November series. The N.-S. Hne, however, on which readings were 

 started the day after the pipe was covered, shows a considerable 

 gradual drop. This was undoubtedly caused by uneven settHng 

 at the ends of the pipes. This was too slow to affect the result 

 for the semi-diurnal period, but made doubtful the results for the 

 fortnightly period. 



The observations and their graphs are given in Tables III and 

 IV and in Figs. 3 and 4. 



In Figs. 5 to 12 these graphs are reproduced on a larger scale, 

 together with those of the calculated values of the readings, multi- 

 pHed by the factor o . 7 for the E.-W. line and 0.5 for the N.-S. Kne. 

 The zero line of the observed readings is corrected to coincide with 

 the calculated curve. 



These curves were divided into periods corresponding to the 

 semi-diurnal lunar tide, 12.42 hours, and the values at correspond- 

 ing intervals of two hours tabulated and divided into groups of 

 ten periods each. 



The results of the comparison of the observed and calculated 

 curves is given in Table II, in which the first column gives the 

 ratio of the observed ampUtudes to the calculated, on the assump- 

 tion of an absolutely rigid earth, and the second the retardation 

 of phase in hours. 



TABLE II 



