26 CV CEARTTERS 
found no explanation of this discrepancy. As a matter of fact the bright 
Bo stars posses a very small proper motion (and hence give a large 
value of R), whereas the fainter stars of this subclass have proper 
motions! exceeding those of all other stars of type B. Is the subelass 
Bo perhaps to be divided into two further subclasses? I can find no 
grounds for such subdivisions in the Harvard Annals. 
11. The investigation made above is, as already mentioned, 
incomplete as far as the stars brighter than the 5" magnitude are 
concerned. Of such stars only those are taken into account for 
which the radial velocity is known. For determining R and M we 
may, however, use all stars for which the apparent proper motions 
in « and d have been determined, whether the radial velocities are 
known or not, Accordingly I have made a new determination of 
U':R, V":R ete, using all stars of the spectral type B brighter 
than the sixth magnitude. 
The results of this investigation are condensed in the table 
Table 4. Corrected values of R and M when all stars brighter than 
the 6" magnitude are taken into account. 
| BE elas Qe. ee AT Buty Un NN, M | PEN: 
Subclass Bo 
0.92937 | — 09.1855 | 09.3480 (125.90) (— 57.55) | 21 
1.4050 | — 0.3615 | 1.4531 | 3.09 | — 2.45 | 44 
<4m.99 | + 0.20207 
<5.99 | + 0.0833 
+ + 
Subclass Bi 
< 4,99 + 0.0765 + 0.4273 = (05002 rt 0.4968 9.03 — 4,78 27 
5.99 + 0.0104 | + 0.5134 — 0.2164 0.5572 | 8.06 214759199 E35 | 
Subelass B2 
<4.99 | — 0.0082 | + 0.6695 | — 0.3379 | 0.7500 5.98 | — 3.88 | 39 
< 5.99 + 0.0117 | + 0.5080 | — 0.4158 | 0.6566 Rey VÄL BE 
Subclass Ds | 
<4.99 | — 0.0359 | + 0.9966 | — 0.7370 | 1.2400 3.62 | — 2.79 |109 
Z 5.99 ‘0.0200 | + 1.0616" | — 0.7167 | 1,2810) | 2:50 39:22] 0 
Subelass B5 | 
«4.99 | — 0.0326 | + 1.1018 | — 0.7272 | 1.3205 3.40 | — 2.66 72 | 
<5.99 | —0.0173 | + 1.0437 | — 0.8840 1.3678 3.98 | — 2.58 | 134 
1 | speak here of reduced proper motions, h. e. of proper motions reduced to an 
apparent magnitude = 0 (the values u and » above). But also the directly observed proper 
motions are rather large, as can be found from table 1. 
