2 Observations

2.1 Photometry

Photoelectric Johnson photometry was carried out on 6 nights in September-December, 1999 using the 0.4-m Cassegrain telescope of Szeged Observatory (Hungary). The detector was a single-channel SSP-5A photoelectric photometer. We made differential photometry with respect to HD 14172 (V= 6 $.\!\!^{\rm m}$96, B-V= +0 $.\!\!^{\rm m}$22, U-B= +0.20, Krzeminski & Serkowski 1967), which is the bright member of the visual double star BDS 1193. The check star was HD 14173 (V= 7 $.\!\!^{\rm m}$20, B-V= +1 $.\!\!^{\rm m}$00, U-B= +0 $.\!\!^{\rm m}$66, Krzeminski & Serkowski 1967), the faint member of the system, located at $\sim$1$^\prime$ from the comparison star. Since the diaphragm of the photometer is 30 $^{\prime\prime}$, we could well measure the individual stars. We note, that U-band observations were carried out only on the last two nights of the observing run.

Figure 1: The light and colour curves phased with the Hipparcos ephemeris. Note the $\sim$-0.1 shift of the maximum indicating the outdated light elements.

 

 

Table 1: The journal of observations.

JD	type	hours	JD	type	hours
2 451 433	BV	3.4	2 452 153	uvby	4.1
2 451 435	BV	3.4	2 452 154	uvby	6.3
2 451 452	BV	5.2	2 452 155	uvby	2.3
2 451 469	UBV	3.3	2 452 156	uvby	5.6
2 451 480	spectr.	4.9	2 452 157	uvby	4.1
2 451 519	UBV	4.4	2 452 158	uvby	5.2
2 451 781	uvby	4.7	2 452 159	uvby	5.4
2 451 785	uvby	1.0	2 452 160	uvby	2.2
2 451 788	uvby	4.0	2 452 161	uvby	6.3
2 451 789	uvby	1.0	2 452 162	uvby	7.3
2 452 151	uvby	4.2	2 452 191	spectr.	2.4
2 452 152	uvby	5.4	2 452 200	spectr.	5.3

The Strömgren uvby photometric observations were acquired on 4 nights in August and September, 2000 and 12 nights in August and September, 2001 using the 0.9-m telescope of the Sierra Nevada Observatory (Spain) equipped with a four-channel spectrograph photometer. The differential photometric data were obtained using the same comparison and check stars (HD 14172: b-y= 0 $.\!\!^{\rm m}$154, m₁= 0 $.\!\!^{\rm m}$111, c₁= 1 $.\!\!^{\rm m}$153; HD 14173: b-y= 0 $.\!\!^{\rm m}$642, m₁= 0 $.\!\!^{\rm m}$257, c₁= 0 $.\!\!^{\rm m}$455, Olsen 1993). The brightness and colour differences of the comparison stars have been found to be constant to $\pm$0 $.\!\!^{\rm m}$01 (suggested by the rms of the data). The overall accuracy of the standard transformations is estimated to be about $\pm$0 $.\!\!^{\rm m}$01 for the V (both from the Johnson and Strömgren data), B-V and b-y data, $\pm$0 $.\!\!^{\rm m}$015 for U-B and m₁ and $\pm$0.02 for the c₁ data. Because of the same comparison, the standardized V light curves from the Johnson and Strömgren measurements are well comparable (the observed light range - 6 $.\!\!^{\rm m}$60-6 $.\!\!^{\rm m}$70 - was the same for both datasets). The journal of observations is presented in Table 1.

We have obtained 3125 individual V points, (512 from UBV and 2613 from uvby), 512 B-V points, 132 U-B points and 2613 Strömgren indices. The whole dataset has been phased with the Hipparcos ephemeris (P= 0 $.\!\!^{\rm d}$1092130, E= 2448500.0700, ESA 1997) and the resulting phase diagrams are plotted in Fig. 1. The observed behaviour of the colour variations is typical in pulsating stars thus excluding the possibility of other type of variation (e.g. eclipsing or ellipsoidal). However, the light curve showed such cycle-to-cycle changes that the assumption of the monoperiodic nature had to be rejected.

2.2 Spectroscopy

Figure 2: Sample spectrum of V784 Cas.

The spectroscopic observations were carried out at the David Dunlap Observatory with the Cassegrain spectrograph attached to the 74 $^{\prime\prime}$ telescope on one night in October, 1999 and two nights in October, 2001. The detector and the spectrograph setup were the same as used by Vinkó et al. (1998). The resolving power ( $\lambda / \Delta \lambda$) was 11 000 and the signal-to-noise ratio reached 70-100. The spectra in 1999 were centered on 6635 Å, in 2001 on 6550 Å and the wavelength span was 200 Å in both cases. The data were reduced with standard tasks in IRAF, including bias removal, flat-fielding, cosmic ray elimination, aperture extraction (with the task doslit) and wavelength calibration. For the latter, two FeAr spectral lamp exposures were used, which were obtained before and after every five stellar exposures. Because of the short period of V784 Cas the observing sequence of FeAr-var-var-var-var-var-FeAr was chosen. Careful linear interpolation between the two comparison spectra was applied in order to take into account the sub-pixel shifts of the five stellar spectra caused by the tracking of the telescope. We chose an exposure time of 6 min, which corresponds to $\sim$0.04 phase of the dominant period. The spectra were normalized to the continuum by fitting a cubic spline, omitting the region of H$\alpha$.

Besides the telluric features on the blue side of the H$\alpha$ line, we could detect a few weak and broad metallic lines. The H$\alpha$ profile remained symmetric during the observations excluding the presence of high-amplitude non-radial oscillations. A sample spectrum taken in 1999 is shown in Fig. 2. We have collected 98 individual spectra for V784 Cas.