32 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Jl 



Table 'K.— Periods Derived from Harmonic Analysis of Trial Periods 



Estimated periods 



First method Second method Third method 



Length 



t.41 

 1.99 

 2.77 

 3.60 

 4.98 

 6.35 

 8.52 

 12. II 

 18.00 



Length 



T.18 

 1.42 

 2.00 

 2.79 



3-6o 

 4.67 

 6.17 

 8.60 



12.20 

 18.06 



Length No. Ampl. 



2.0 



3.5 



4.8 



6.0 

 9.0 



1-7 

 1-3 

 I.I 

 i.o 



True 

 azi- 

 muth 



180° 



50 



20 



230 



270 



75 

 315 



42 

 330 

 220 



In obtaining- the results given under the heading " First Method " 

 in table X every possible period indicated by the plots was taken and 

 then a curve formed show^ing the frequencies of periods betw^een 

 intervals of half-days. From the maxima of frequencies the prob- 

 able true periods v^ere selected and the means of the indicated periods 

 near these maxima v^ere obtained. The number of cases are given 

 in the table. 



In obtaining the results indicated under " Second Method " only 

 the periods running through or near the points corresponding to the 

 maximum ampHtudes were used. This method corresponded most 

 nearly with the instructions prepared in advance and evidently gives 

 the best results. 



