192 G. H. KNIBBS. 



A count was then made of the distribution of the poles of 

 proper motion over the celestial sphere, and after shewing that 

 the approximate mean error is sensibly identical for each class, 

 from which it appears that the influence of the chance errors of 

 observation are unrecognisable, Kobold deduced the coordinates 

 of the sun's motion from the stars in each class as follows: — 



Class 



Stars. 



«(+) 



R.A. 



D. 





I. 



24 



<:0-°6 



264-°4 



+ 12-°0 



Epoch 1810-0 



II. 



43 



0-°7 to l-°0 



264-3 



-3-5 





III. 



101 



1-1 „ 18 



2641 



-0-7 





IV. 



210 



1-9 „ 3-2 



265-0 



-4-7 





V. 



386 



3-3 „ 5-8 



267-4 



-1-4 





VI. 



636 



5-9 ,,10-5 



267 3 



-4-3 





Total 1,400 Mean result 266-6 - 3-0 



Treating the whole of the equations the result was 

 R.A. = 266-°5, D.= -3°1 

 practically identical with the mean result as shewn. 



Kobold discussed this with respect to the quantity of the solar 

 P.M., from which it appears that it is considerably smaller for the 

 declination — 3° than for +31°, the former assumption giving 



V = 0'61 German miles per sec, 

 while the latter gives 1*23. He moreover deduced the quantity 

 of the sun's motion in relation to the members of two series of 

 stars of which the parallaxes are known, and also the velocity in 

 the line of sight from spectroscopic observations. The first series 

 contains 11 stars, the second 18. The general result is that the 

 evidence points to the apex being very near the celestial equator 

 rather than about 30° away. 



(100) Gylde'n, 1894- (Aug.) — The relations of magnitude and 

 proper motion to parallax are obviously important in connection 

 with the analysis of the sun's motion. An attempt was made by 

 Gylden in 1894 by discussing stars of known parallax with various 

 proper motions and magnitudes to define those relations quantita- 



