3 New photographic magnitudes from Harvard and Sonneberg sky patrol plates

3.1 Comparison star sequences

For our aim to determine the long-term behaviour of the light variations, brightness estimates by eye are the only appropriate way to cope with the large numbers of plates. The accuracy is expected to be sufficiently high to detect Algol-like minima with amplitudes $\ge0.5$ mag. To have a reliable basis for the brightness estimates and to limit the photometric errors, we determined the magnitudes of a large number of comparison stars. The necessary observations were done with a CCD camera attached to the 60/90/180 cm Schmidt telescope of the Universitäts-Sternwarte Jena at Großschwabhausen Observing Station. As primary standards we used the stars observed by Pugach & Kovalchuk (1983). The only omission was their comparison star "g'' for BH Cep. This star is very red and the authors consider their magnitude measurement very uncertain. Therefore we discarded this star as second standard star for the photometric sequence for BH Cep and used comparison star "$\alpha$'' from Wenzel & Brückner (1978) instead. Reduction of our photometric data gives $V=11\hbox{$.\!\!^{\rm m}$ }66$ for Pugach & Kovalchuk's star "g''. (The very same star is denoted by "i'' in Table 2.) The colour system of our CCD camera is virtually identical with the standard UBV system. Therefore, no colour transformations have been applied.
Typical values for the rms errors of the magnitudes derived by our CCD photometry are $\Delta m \approx \pm 0.02$ mag for , $\Delta m \approx \pm 0.03$ mag for , and $\Delta m \approx \pm 0.10$ mag for .
In Fig. 1 the comparison stars for VX Cas, BH Cep, BO Cep, and SV Cep are identified. In the case of SV Cep we present an updated version of our earlier identification map (Friedemann et al. 1992). To avoid confusion, we left the labelling of the comparison stars unchanged but added two more comparison stars. Table 2 gives the B and Vmagnitudes measured by us. Since the B magnitudes are very similar to photographic (blue) magnitudes, they were used with the visual estimates on the photographic plates without any further transformations.

For an identification map of the comparison stars for the variable RZ Psc and their magnitudes see Friedemann et al. (1995).

   

Table 2: B and V magnitudes of the comparison stars

VX Cas				BH Cep				BO Cep				SV Cep
Star	B	V		Star	B	V		Star	B	V		Star	B	V
a	10.38	9.31		a	9.95	-		a	8.87	-		r	10.11	9.74
b	10.57	9.63		b	10.52	10.12		b	9.19	-		s	10.76	10.34
c	11.31	11.03		c	11.40	11.08		ca	11.60	10.94		bb	10.94	10.73
d	11.42	11.03		d	11.66	11.11		d	11.68	10.10		c	11.54	11.10
e	11.44	11.21		e	12.12	11.52		e	12.03	11.44		t	11.90	11.66
f	11.80	10.83		f	12.16	11.59		f	12.14	10.83		f	12.55	12.17
g	11.81	11.55		g	12.58	11.93		g	12.63	12.06		e	12.80	11.66
h	12.13	11.48		h	12.77	11.91		h	12.83	11.66		g	12.82	12.36
i	12.22	11.43		i	13.08	11.66		i	12.85	12.14		i	13.05	12.54
k	12.67	12.26		k	13.42	12.27		k	12.88	11.84		d	13.28	12.77
m	12.71	10.93		m	13.65	12.84		m	13.23	12.80		h	13.58	12.63
n	12.74	11.39		n	13.87	13.11		n	13.39	11.67		k	13.88	13.30
o	12.75	12.00		o	13.95	13.23		o	13.42	12.74		m	14.48	13.96
p	12.96	12.54		p	14.20	13.27		p	13.44	12.20
q	13.20	12.81		q	14.26	13.39		q	13.78	12.99
r	13.25	12.45		r	14.34	12.64		r	13.90	12.67
s	13.56	12.92		s	14.54	13.69		s	14.04	12.97
t	13.58	12.85						t	14.26	13.62
u	13.70	12.41						u	14.27	12.94
v	14.17	13.26						v	14.33	13.61
								w	14.52	13.26
								x	14.62	13.03

$\parbox{18cm}{\noindent \\ $^a$\space Close double star.\\ $^b$\space Possibly variable. The star brightened up sporadically by about 0.5~mag.}$

3.2 Accuracy of the magnitude estimates

Within the long time interval covered by our work, many different sorts of photographic emulsions were used in the sky patrol work. In order to cope with this diversity, we restricted ourselves to blue sensitive photographic plates. Systematic differences were minimized by using the same sequences of comparison stars on all plates. The accuracy of the brightness estimates depends on the local sensitivity fluctuations of the emulsions and the accuracy of the estimates themselves. The individual step sizes among the authors varied from 0.08 up to 0.15 mag. Thus, the uncertainty of the magnitudes should not exceed $\Delta m \approx \pm0.3$ mag in most cases. This is in accordance with the statistical error which we derived when several plates per night were available.

A feature common to all our data is the marked horizontal striation in the lightcurves (see Fig. 2). This striation grows to prominent voids in some regions of the lightcurves of WW Vul, BH Cep, and SV Cep. The fine striation is a natural consequence of the fact that due to the limited accuracy of the brightness estimation by eye the magnitudes are assigned in discrete steps to the variable. The broader horizontal voids of missing data points have an additional cause. As already discussed with the presentation of the data for WW Vul (Friedemann et al. 1994b), statistical analyses revealed that some observers were evidently reluctant to establish equal brightness for the variable and the comparison star. As a consequence data points which represent the magnitudes of the comparison stars are small in number or absent at all. The vertical width of the voids does not exceed 2 steps and, therefore, the maximum error estimated in the last paragraph is not increased by this personal bias.

No systematic differences were found among the brightness estimates by different observers in overlapping regions of the lightcurves. There are also no systematic differences between the eye estimates and the photoelectric measurements, as the inclusion of the latter in Fig. 2 shows. Of course, we must keep in mind that the accuracy of the magnitudes derived by visual brightness estimates is very low if they are compared with photoelectric data. Nevertheless, brightness estimates remain the only effective way to provide data for statistical analyses of lightcurves as long as the information of the plates is not yet available in digitized form.

   

Table 3: Basic information on the new data sets

Star	Period	$B_{\rm max}$	$B_{\rm min}$	Archive	N	$\Sigma$	Authors
	(JD)	($^{\rm m}$)	($^{\rm m}$)
VX Cas	2414633-2439917	10.5	12.6	Harvard	1817		Gü
	2436426-2450370			Sonneberg	1020	2837	Ru
RZ Psc	2414631-2439141	10.9	13.8	Harvard	964		Rei
	2425510-2448983			Sonneberg	1022a	1986	Home
BH Cep	2414673-2439692	11.2	12.2	Harvard	2018		Fr
	2436019-2450463			Sonneberg	1958	3976	Spl
BO Cep	2414673-2439692	11.7	13.3	Harvard	1702		Fr
	2436019-2450463			Sonneberg	1977	3679	Spl
SV Cep	2414673-2447857	9.7	12.7	Harvard	2232		Fr, Rei
	2425503-2450463			Sonneberg	2298	4530	Fr, Gü, Spl

$\parbox{18cm}{\noindent \\ $^a$\space Most of these data have been already published by \cite[Home et al. (1994)]{hom94}.}$

3.3 The new data

Table 3 provides basic information on the collected data, such as the time intervals covered, the number of useful plates, and the name of the person who did the estimates. The magnitudes, which cover the period from 1898 to 1997 in most cases, are available in electronic form from the Centre des Données Stellaires at Strasbourg, France.