The general Characteristics of the frequencyfunotion of stellar movements 



7 



called q . As a working hypothesis I assumed as has done Charlibb, that those 

 parameters have the same value for the whole heavens. This assumption is ne- 

 cessarily very provisional. It is for instance exceedingly probable that the mean 

 parallax varies with galactic latitude, though as we here only use stars brighter 

 than the magnitude 6.0 the variation cannot be very great. However, I have been 

 able, when studying the results, to let the working hypothesis drop, having procured a 

 method to correct the final results for any systematical variation of the constants. The 

 amount of variation of the parameter g' in different parts of the sky is very difficult 

 to obtain, evidently the time is not yet ripe for such a determination. Besides it 

 is in the present case of only slight importance. The variation of the mean parallax, 

 however, I have tried to determine by comparing the mean parallactic motion of 

 the several regions in the sky with the mean parallactic motion of the whole heavens. 

 The correlation coefficients and the higher characteristics being abstract numbers 

 (of dimension zero) they are practically independent of the variation. Only the axes 

 of the correlation ellipsoid (mean internal motion) are affected and duly corrected. 



