390 E. W. Brown — Apparatus for Tidal Analysis. 



18 X 6 + 35 X 5 = 283). We have thus the series 6, 5, 5, 

 repeated 17 times, then 6, 5, and finally 6, 5, 5, to the end. 

 The above examination of the first eleven days shows that the 

 first no-step day is day 5, the next day 11, which facts give the 

 start of the 6, 5, 5 series. It is easy to run through and find 

 the days on which h comes into the first column. Thus the 

 instructions to the operator are as follows : 



' Place each band one step to the left of the previous band 

 except on the following days when no step is to be made : 

 5, 11, 16, 21, and thereafter the 6th, 5th, 5th in succession until 

 the no-step day 277 is reached. The days 283, 288, 294, 299, 

 304 are no-step days and thereafter the 6th, 5th, 5th in succes- 

 sion until the end. For testing the arrangement, note that 

 the observations at h for the following days are in the same 

 column :— 0, 29, 58, 59, 88, 118, 147, 177, 206, 236, 263, 295, 

 324, 354.' 



Similar instructions may be made out for the other periods. 



This apparatus is more convenient for tidal analysis than 

 that which I described* some years ago and which has been 

 continuously in use for the synthesis of the small terms in the 

 moon's motion which are to be inserted in the lunar tables. It 

 is not, however, applicable, like the earlier device, to general 

 harmonic analysis. 



Yale University, New Haven, Conn., 

 Feb. 8, 1915. 



* Monthly Notices E. A. S., vol. lxxii, pp. 454-463. 



