COMMISSION 27 OF THE I. A. U.
       INFORMATION BULLETIN ON VARIABLE STARS
                    Number 1023

                                 Konkoly Observatory
                                 Budapest
                                 1975 July 23




      A CRITICAL REMARK ABOUT THE DETERMINATION OF
                   MINIMUM TIMES	


While investigating the period changes of the W UMa star 44i Boo,
I looked through more than 25 years of photoelectric mininum time
determinations. A closer inspection of the material revealed an
increasing scatter in the more recent O-C data, and it seems very
likely that this effect is not an intrinsic property of the star,
whose small amplitude (Delta m ~~ 0.15) makes it a difficult object, but
originates from a changing approach on the observer's part.

Fig. 1 illustrates this effect. The data have been compiled from
various sources (Schneller 1965, Svechnikov and Surkova 1973, IBVS
530, 647, 789, 889, and 937). The following light elements, which
satisfy the observations very well, have been used for the 
determination of the O-C values:

    2438513.4166 + 0.2678143 * E (1947 - 1967) (Pohl 1967)
    2439852.4903 + 0.2678159 * E (1968 - 1973)

Each point of the diagram gives the standard deviation, derived
from five O-C values, which are so closely spaced that the period
variation during this time interval can be neglected. Around 1964,
the scatter of the data increased from about 0.0009 to 0.0022 days.

A possible explanation of this general trend is, that the early
observers were interested in the complete light curve and measured
always large parts of it, while the more recent observations are
"only" used for minimum time determinations, with probably short
observing runs and only a few data points for each minimum.

The accuracy of 0.0022d equals that of a visually determined minimum
time of a typical Algol variable (Duerbeck 1975), and Herczeg and
Frieboes-Conde (1979) have shown that this accuracy is not sufficient
to draw any conclusion about the character of the period changes.

I therefore strongly request that the observers of minimum times
1. increase the accuracy by using more observations, and
2. apply methods for the determination of minimum times which include
  the derivation of errors (e.g., Kwee and van Woerden 1956, Breinhorst 
  et al. 1973), so that the compiler can judge the quality of
  the material.

                        [FIGURE 1]

  Fig. 1. Standard deviation of the O-C values for 44i Boo.



                       H.W. DUERBECK
                    Observatorium Hoher List
                 der Universitaets-Sternwarte Bonn
                    5568 Daun/Eifel, F.R.G.



References

Breinhorst, R.A., Pfleiderer, J., Reinhardt, M., Karimie, M.T.,
  1973, Astr. Astrophys., 22, 239. [BIBCODE 1973A&A....22..239B ]
Duerbeck, H., 1975, to be published.
Herczeg, T., Frieboes-Conde, H., 1974, Astr. Astrophys., 30, 259. [BIBCODE 1974A&A....30..259H ]
Kwee, K.K., van Woerden, H., 1956, Bull. Astr. Inst. Netherl., 12, 327. [BIBCODE 1956BAN....12..327K ]
Pohl, E., 1967, IBVS 209.
Schneller, H., 1965, Astr. nachr. 288, 183. [BIBCODE 1965AN....288..183S ]
Svechnikov, M.A., Surkova, L.P., 1973, Perem. Zvezdy 18, 571. [BIBCODE 1973PZ.....18..571S ]