Commission 27 of the I. A. U.
            Information Bulletin on Variable Stars
                          Number 5





                                             Konkoly Observatory
                                             Budapest
                                             15 January 1962


        A WARNING TO THE OBSERVERS OF BETA CMa-TYPE STARS


     It is the current policy at most observatories to restrict the telescope 
time available for the observation of the (all bright) Beta CMa stars
to the bright moon period, when work on fainter objects becomes hard or
impossible.
     Observers should be well aware of this serious danger hidden in
this apparently sound procedure, at least when their observations are
aimed at the determination of period lengths and amplitudes in these
multiperiodical stars.
     Let us indeed first consider a short period variable with a single
period P and suppose that the observations on it lie clustered around
epochs (full moon) that are separated by a characteristic interval T.
Let be P=T/n, with n large, but not necessarily an integer. In the
course of a periodogram analysis the exact period will of course
yield the global mean light or velocity curve with the largest amplitude
possible. But the periods P=T/n+i (i 1,2,3,...) will also fit the observations 
on the individual full moon periods quite well (as long as i<<n) and
since the mean curves derived for the latter are all in phase, the global
mean curve will have an amplitude that decreases but slowly with 
increasing i. On the contrary, for i=1/2, 3/2, 5/2,... the various groups
are of opposite phases and the global mean curve will never
reach a significant amplitude. Consequently the periodogram will show, instead



of its normal aspect, a large number of narrow bands closely packed
together and having heights that go increasing towards the central one,
which corresponds to the true oscillation both in period and amplitude.
The latter thus remains easily recognizable in this case.
     But if we consider next a variable in which several true periods lie
closely together, as is the case of the Beta CMa stars, the band patterns
from the various periods will interfere and produce a very misleading
spectrum; several of the resulting bands will show high peaks, while
others will have their heights reduced almost to nothing. The identification 
of the physically meaningful bands will then become a very tough job.
If our general knowledge of these stars may prove a competent guide at
this stage, it is of little help at the next: the determination of accurate
periods and of reliable amplitudes. Both remain unknown.
     May both the observers and those who have to allocate telescope
time keep in mind this special aspect of the investigation of stars with
multiple short periods.



                                      A. van HOOF
                                 University of Louvain
                                        Belgium