( '-^l-^ ) 



1 coiiiinenced iny calculations hv deducing also for the period 

 1.S90.0 to 1899.8 tlie value of the vearly c-term, found afterwards by 

 KiMURA, as far as the published observations would allow it. The 

 separate results came out with considerable uncertainty, but takinji,- 

 the mean of the values, found for : for the corresponding tenth 

 parts of the different years, 1 arrived at 10 mean values, which 

 were represented as well as possible by the following' siiuisoïd ; 

 ^ = + 0"0043 sm If' + O"0221 cos if 



in which if? is being counted from the beginning of the year. 



From the values of Ay, corrected for this j-term, I computed 

 the rectangular polar co-ordinates ./■ and // foi' that period, so (hat I 

 obtained a continuous, and homogeneously reduced series of these 

 co-ordinates, from 1890.0 up to 1908.5. From this 1 deduced in 

 lirst approximation the elements of the yeaily component for (hiee 



periods : 



a from 1890.0 to 1897.0 



b from 1895.5 to 1902.5 



c from 1901.0 to 1908.0 



takin^'- each time 7 years together in oi-der to eliminate the second 

 com[)onent, whose period was thus in this tirst approximation su[)- 

 posed to be exactly 14 months. 



For these three periods 1 found a yearly ellipse, showing slight 

 variations in size and shape, and in the position of the axes. This 

 need not to be a cause of wonder, however, when we accept clianges 

 of mainly atmospheric nature, e. g. varying dislribulion of atmospheric 

 pressure, accumulation of snow and ice in winter, as the chief factor 

 in l)ringing about this component. 



A single result of this investigation deserves to be shortly men- 

 tioned here. 



While for the periods a and b the zero-values of ^ and ?/ proved, 

 that the adopted origin of co-ordinates coincided practically exactly 

 with the real mean pole, I found in this first approximation for the 

 central co-ordinates in the period c : 



^ — + 0" U09 n= + 0".032. 



These values indicate, that for the later years the adopted origin 

 deviates sensibly from tiie niean pole. The observations of this period 

 have all been made at the six international latitude stations, and 

 have been reduced uniformly by Prof. Ai.bhecht in his Ri'sultate 

 Bnd. I, II and 111. The origin of co-ordinates chosen by him, and 

 adopted as "mean pole", coincides fairly accurately with the centre 

 of the orbit of the pole resulting from the observations from 1899.9 



