COMMISSION 27 OF THE I. A. U.
          INFORMATION BULLETIN ON VARIABLE STARS
                       NUMBER 10

 
                                       Konkoly Observatory
                                       Budapest
                                       18 May 1962

            MULTIPLE PERIODS IN 15 CANIS MAIORIS

     According to C. R. Lynds, J. Sahade and O. Struve (Ap. J.,
124, 321 p, 1956) the radial velocity variation of 15 CMa obeys a main
period of 0d.184558 (2K=6.5 km/sec). Their computed mean RV
curve reveals, however, in a striking manner the existence of a
secondary period of exactly half that length and with a corresponding
amplitude of 2.5 km/sec.
     Such an exact 2:1 ratio has been found to exist between the
periods P_{1,3} and P_2 (s. foot-note+) in each of the stars Nu Eri (1) and
Beta Cep (2), which are the immediate neighbours of 15 CMa when the
Beta CMa stars are listed according to period length. Taking into
account the similarity in behaviour of the group members, the two
periods discovered by Lynds et al. must be considered as the star's
P_{1,3} and P_2 respectively.
     Postulating further that the ratios

     P_{1,3)} = 1.01*P_0 ; P_1 = 0.675*P_0 ; P_2 = 0.505*P_0 etc.



     + The symbol P_i is used to denote the free pulsation in the
fundamental mode (i=o), the first overtone (i=1), the second overtone
(i=2) etc., while the symbol P_{i,j} denotes the difference oscillation
between P_i and P_j.

     (1 ) A. van Hoof, Zeitschr. f. Astroph., 53 ,106 S., 1961
     (2)       id          ibid.                (in press)


                             - 2 -

found in both Nu Eri and Beta Cep hold also good for the present star,
we are able to compute the various other periods that must be expected
in 15 CMa. In this manner we found:

P_0 = 0d.182586            (6.0 km/sec)   P_{0,2}= 0d.186663   (5.3 km/sec)
P_1 =   .123154            (2.5   "   )   P_{1,3}=   .184558   (6.5   "   )
P_2 =   .092279            (2.0   "   )   P_{2,4}=   .182586 = P_0
P_3 =   .074606            (0.3   "   )
P_4 =   .0620115           (1.5   "   )

The amplitudes added are those of the mean curves computed from
Lynds' Sahade's and Struve's observations for each of the periods listed.
It should be stressed, however, that this material is quite
inadequate for the determination of the amplitudes of the longer periods
P_0, P_{1,3} etc. This material has been collected in three groups of three
consecutive nights and the mean epochs of these groups are separated
by intervals which by accident correspond roughly to an integer number
of each of the periods P_0, P_{0,2} and P_{1,3}, so that the phase relation of
these three oscillations is practically the same in the three series of
nights. Consequently the mean RV curves computed with these three
periods must practically have the same amplitude, which is simply the
resultant amplitude of the superposition in a particular phase relation
of the three component curves, the amplitudes of which remain unknown.

      It is further obvious that the scantiness of the observations
does not permit the accurate determination of any period or amplitude.

      With this restriction in mind we want nevertheless to mention
the weakness of the oscillation P_3. The same weakness was noticed
earlier in Nu Eri and in Beta Cep.

                                                A. VAN HOOF
                                           University of Louvain
                                                  Belgium