( 216) 



5 = +U".OUG ./•., = r^ ,in (ip^ + 184°.0)) 



.... (4) 

 rt= ^ O .040 //, = c, cos (V^, + 184°.0) j ^ ' 



where again »f\^ lias been counted from 1907.5, and furtlier: 



from 1904.0 to 1907.0: c, = i)"A18 



tVoni 1908.0 to 19J1.0: c, = .250. 



From ["2) and (4) 1 now i*om|nited, according- to (1), the (lieoretical 

 values of .6' and //, and compared them with At, hrkcmt's results from 

 ihe observations. Tal)le I gives the i-esnlt of this comparison, the 

 adopted unit being Wo^,^, second of arc. 



The differences — C (Observation — Computation) give as mean 

 enoi' of my computed values: 



J904— 1907 : ///, = ± 0".0204 



1908—1911 : db .0237 



while, as I said before, the mean error of an observed co-ordinate 

 is estimated by Alrkkcht at =b 0".020. 



I think, that from this I may conclude, that within the admissible 

 limits of error of the observations the polar motion is represented 

 by a yearly and a J4-monlhly component, as expressed by the formulae 

 (2) and ^4). 



From this it follows : 



'J. that a change iji the 14-nionthly motion must have taken place 

 in the course of the year 1907 ; 



2. that the perturbation did not cause any appreciable change of 

 phase, or displacement of the mean pole ; 



3. that the change in the motion is wholly owing to a gradual, 

 or more or less sudden increase of the amplitude of the 14-monthly 

 motion from 0".12 to 0'.25. 



In the following manner 1 have attempted to determine more 

 precisely the very moment and the nature of the perturbation. 



From the formulae found for the period 1904 — 1907 1 deduced 

 the co-ordinates for 1907.0 — 1908.0, such as they ought to have 

 been, had the motion of the i)ole remained undisturbed, and com- 

 pared them with the observed positions. The numerical results of 

 this comparison are contained in Table II. 



The columns — C yield as mean error ± 0".0461, which is far 

 more than the accuracy of the observations allows for. Moreover a 

 glance at these differences makes it clear, that although the agreement 

 up to 1907.3 or 1907.4 may be satisfactory, the differences found 

 afterwards are decidedly not admissible. 



Secondly 1 compared the observed co-ordinates with those which 



