27 37 S 



parallax of the Plejades is calculated to be "'0147 according to the proper motion 

 of Boss or "'0119 after having applied the Kapteyn correction. 



If the relative proper motions of the Pleiades (also some unpublished material 

 was used for this purpose) are plotted separatelj' for each magnitude, it is seen 

 that the stars of the 4 or 5 brightest magnitudes nearlj' all belong to the system. 

 The diagram containing the stars between the photographic magnitudes 10"^ and 

 11™ still markedly shows that a group of these stars is moving with the cluster, 

 while there is no more evidence of this when we consider stars between 11™ and 12"'. 

 Regarding still fainter stars there are hardly more objects showing a proper motion 

 common to that of the Pleiades, than should be expected by accident. The good 

 agreement of the colours of such fainter stars, apparently sharing the motion of 

 the cluster, with those expected, is deceiving, because these colours wall also be 

 about normal for faint stars of such proper motions which are not members of the 

 system. It is thus very probable that R. Trümpler (Lick Bull. 333, Vol. 10, 110; 

 1921) has overestimated the number of faint stars belonging physically to the group. 



I have therefore found it safest, until more accurate proper motions are avail- 

 able, to confine myself to stars brighter than 11™ photographicallj' when trying to 

 separate the members of the system from the rest. A list of 66 stars thus considered 

 as physically belonging to the group of the Pleiades is given in Table 14. The 

 most remarkable fact shown by this table is the very regular change in colour 

 with the magnitude. Between 4™'5 and 11™ the relation between /; and Wpg is 

 nearly linear. A least square solution gave for these stars the formula 



/;= —-662 + -0945 mpg 



The deviations — C in /; are given in the fourth column. Their smallness is 

 a sign of the accuracj' of the /; -values. As there may be a slight preponderance of 

 positive values of — C in the neighbourhood of 8™ I have compared the — C 

 values of two consecutive stars in the magnitude sequence in order to determine 

 the accuracy with which the colour or // value of a star maj^ be predicted by its 

 magnitude. Taking the mean error of the determination of /; into account, it is 

 found that the mean "error" or deviation of /; from the normal value for a given 

 magnitude is i ™028. To that degree therefore members of the Pleiades of the 

 same magnitude have the same colour index. It is to be remembered, that double 

 stars counted as single and consisting of two nearly equal components should give 

 — C values about -H™'09 greater than for single stars. 



The 10 stars brighter than 5™'5 are all practically of the same colour, the 

 deviations from the mean Ii = — '20 not being greater than the observational errors 

 zb™'015. 



For the physical members of the group of the Hyades I have formerly (A. N. 

 5000, Bd. 209, 120; 1919) found the relation /— "364 = -0935 (m^j — 6-596). The 

 coefficient JljJnipg is thus practically the same, viz. +'094, for the Hyades and 



49* 



