COMMISSION 27 OF THE I. A. U.
        INFORMATION BULLETIN ON VARIABLE STARS
                     Number 1318

                                 Konkoly Observatory
                                 Budapest
                                 1977 August 8





     A RECONSIDERATION OF THE ORBITAL PERIOD OF AZ Cas



   Ashbrook (1956) found that the cool supergiant AZ Cas was
an eclipsing variable with a period of about 9.3 years. From
magnitudes estimated on Harvard patrol plates covering the
eclipses between 1901 and 1947 he derived the ephemeris:
Min = JD 2432484 + 3406 E.
   Larsson-Leander (1959) noted that his photoelectric 
coverage of the 3rd and 4th contacts during the 1957 eclipse 
occurred about 6 days earlier than predicted from Ashbrook's 
ephemeris. In a discussion of the eclipse times Bonnell and
Herczeg (1976) have concluded that the period should be revised
to 3402d. However, their ephemeris predicted the 3rd and 4th
contact of the 1975 eclipse to occur on 1975 Nov. 17 and 27+-1d
respectively. Florkowski's (1975) photoelectric coverage of the
eclipse shows the 4th contact occurred on 1975 Dec. 8, 
suggesting that P = 3402d is too short. On the other hand, 1st and 2nd
contacts as given by Florkowski (1975 Aug. 2-3) and Tempesti
(1975) (1975 Aug. 11) occurred 10 days too early to satisfy
Ashbrook's ephemeris. Using all of these data we have derived
periods based on comparing times of 1st, 2nd, 3rd and 4th 
contacts to be 3403+-1.2,3402+-1.8,3406+-0.7 and 3405+-1.4, respectively. 
Each figure is derived from between three and five eclipses
between 1901 and 1975, but because of incomplete coverage not
all the periods are determined from the same eclipses. From all
of these data we derive an ephemeris:
           Min = JD 2432481 + 3404 (+- 1d) E.
We note that there is good evidence that in the last two eclipses
for which we have observations (1957 and 1975) the duration of
the eclipse was longer by ~9 and ~20 days than the mean  duration
 (~86d) of the eclipses observed by Ashbrook. This change in
eclipse duration does not extrapolate back through Ashbrook's
observations, however. The effect is to make the periods derived from 
1st and 2nd contacts shorter than the mean, and similarly those 
periods determined from 3rd and 4th contacts somewhat 
longer. The mean value should be unaffected. The increased
duration of the minimum may have caused Bonnell and Herczeg to
suggest an abrupt period change after the 1947 eclipse.
    The same value of the period (3404d) was also derived by
a slightly different method, using the deviations of times of
mid-eclipse from the values predicted by Ashbrook's ephemeris.
This method is not quite as sensitive since the whole of totality
was not usually observed and one must assume something about the
eclipse duration, which appears to be variable. Presumably the
variation in eclipse duration may be due to a change in radius
of the primary star. Spectroscopic data on AZ Cas are discussed
in full elsewhere (Cowley, Hutchings, and Popper 1977).
    We thank Dr. Ashbrook (1975) for the use of his eclipse data
in advance of publication.





                   A.P. COWLEY and J.B. HUTCHINGS
                   Dominion Astrophysical Observatory
                   Victoria, Canada





References:
Ashbrook, J. 1956 Harvard Obs. Announcement Card No. 1340
Ashbrook, J. 1975 Private communication
Bonnell, J.T., and Herczeg, T.J. 1976 IBVS No. 1146
Cowley, A.P., Hutchings, J.B., and Popper, D.M. 1977, in press
Florkowski, D.R. 1975 Private communication
Larsson-Leander, G. 1959 Arkiv Astron. 2, 347 [BIBCODE 1960ArA.....2..347L ]
Tempesti, P. 1975 IAU Circ, No. 2825 [BIBCODE 1975IAUC.2825....3T ]