( 526 ) 

 Apparent vertices (convergents) and stream-velocities. 



exclusively to rely on the data for the G- and A-stars. For the tirst 

 stream the data of the A-stars may practically be neglected ').'' The 

 reason is that not only the elements of the A-stars are so much 

 more uncertain, but also that the amount of the velocity both relative 

 to the G- and to the H-stars is so small. 



The only somewhat reliable results therefore are as follows ^) : 



stars stream relat, veloc. inclination on 



What is particularly remarkable in these numl)ers is, that the 

 inclinations on the JMilky way are so small. 



Former investigations have shown that the motion of the two 

 streams of G-stars relative to the centre of gravity of all the stars 

 is perfectly parallel to the plane of the JMilky way '). In my first 



1) The numbers on this line are those of Eudington (Monllil. Not. Nov. 1910). 

 In reality E. has treated together the stars of the spectres A F G M K. The mean 

 type is G and I have considered E's results as being valid for this type. E's 

 il3sults for velocity were changed into kilometers by the aid of Campbell's last 

 determination of the sun's velocity = 19.5 kil. (Lick. Bull. W. 196). 



2) In reality only the early A-stars i. e. only the spectral classes AjAjAjAs. 



3) They are of course particularly important for settling the question whether 

 or not the change of the elements with fime is really gradual. 



•*) For the benefit of Ihorie, who, in spite of their great uncertainty, might 

 wish to know the corresponding results for the A-stars of stream I, I give the 

 following numbers. 1 will attach a sign to the inclination in order to show that 

 the two relative velocities deviate from the Milky way towards different sides. 

 stars stream relat, veloc. inclin. on Milky way 



G and A I 7.0 kil. + 22°7 



A „ B 1 5.9 „ — lOH 



°) In other words that the relative velocity of the two streams of G-stars is 

 parallel to the Milky way. 



