Although they are all standard of the Geneva system, we checked the absolute photometry of each comparison star in order to detect possible spurious phenomena. All were perfectly stable, with the exception of HD 190422, comparison star of HD 191439 and HD 192674, which was found to be slightly variable with a period of about 3 days. With a spectral type F8V, HD 190422 may be a small-amplitude cepheid. A more detailed analysis using additional data is being carried out.
The absolute data of HD 191439 and HD 192674 were analyzed instead
of the differential ones. Variations could be detected in
the former, with a period of three days, close to that
of HD 190422. Because of this
coincidence, a blind
analysis of the differential data would not have
raised any suspicion about the comparison.
The period of HD 192674 is less clear. We favor the value 
.
Another coincidence jumps to the eyes. It was immediately apparent
during the analysis that an uncomfortably large number
of stars had similar periodograms,
with peaks at the same positions (see Fig. 1 (click here)). It is well
known that the distribution of the periods of
Ap stars has a maximum between 2 and 2.5 days, but it was hard to
believe that three stars in such a small sample
could have almost
exactly the same period of 216!
We initially suspected some
flaw in the measurement or reduction procedures.
However, other stars measured in almost exactly the same conditions
(comparisons, or nearby Ap stars) did not show the same behaviour.
Moreover, the phase diagrams constructed with these periods
were quite different for the various stars
(see Fig. 2 (click here)).
We were then forced to admit the possibility of those coincidences.
Aside from these three stars, a fourth one has a slightly
longer one, and a fifth one has almost the double period.
Looking
at older data, we found that coincidences abound in Ap stars
studies.
In an earlier study of 56 stars,
Mathys & Manfroid (1985) found three stars with a period within 
of 
and two stars with the longer period 
.
Figure 2: Phase diagrams for the differential data
of HD 8717, 200623 and 207259 in B1-V1. Time origin
is JD 2 450 000 0.
The curves are second-order Fourier series. These stars
have the peculiarity of showing almost exactly
the same period 
(see Fig. 1 (click here)).
Tick marks on the vertical axis are separated by 0.01 mag.
The complete set of lightcurves is available at the CDS
Among the Ap stars classified as uncertain by Renson (1991), HD 8717 shows Ap-like variations, which would confirm its status in this category.
The most uncertain cases in Table 2 (click here) are HD 192674 and HD 192687. The proposed periods are only tentative. There are no indication of variations in the suspected Ap stars 200199 and 204367.
HD 191439 and HD 200623 had already been observed by Vogt & Faundez (1979). From four uvby measurements these authors could not recognize the variability of both stars.
Inspection of the lightcurves, and Table 2 (click here), show that, despite the weaker signals, magnitudes and color indices involving narrow-band filters often provide cleaner light curves, a conclusion already reached from the comparison of Strömgren's and Johnson's photometries. In this respect, the neglected B1-V1 index proves to be interesting. We also note that the largest amplitudes often occur in the U band, which is a well-known characteristics of Ap stars.
Acknowledgements
We are grateful to the Geneva Observatory and especially to Pr. G. Burki for making possible these observations at the Swiss telescope on La Silla. Our thanks also go to Mr. C. Richard and Mr. B. Pernier who carried out the data reductions.